(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Dynamos and dynamo design ; Direct current motors ; Alternating currents ; Alternators ; Alternating-current apparatus"

This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project 
to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject 
to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books 
are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover. 

Marks, notations and other marginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 
publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing this resource, we have taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 

We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain from automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attribution The Google "watermark" you see on each file is essential for informing people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright infringement liability can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 

at http : //books . google . com/| 




Dynamos and dynamo design; 
direct current motors; ... 

International Correspondence Schools 




K.R WENDT LIBRARY 
UW COLLEGE OF ENGR. 

215 N. PAMHAM AX/rMllg 

MADl 




Digitized by VjOOQIC 



Digitized by VjOOQIC 



( 



Digitized by Google ^ 



Digitized by VjOOQIC 



Digitized by VjOOQIC 



Digitized by VjOOQIC 



Digitized by VjOOQIC 



INTERNATIONAL 
LIBRARY OF TECHNOLOGY 



A SERIES OF TEXTBOOKS FOR PERSONS ENGAGED IN THE ENGINEERING 

PROFESSIONS AND TRADES OR FOR THOSE WHO DESIRE 

INFORMATION CONCERNING THEM. FULLY ILLUSTRATED 

AND CONTAINING NUMEROUS PRACTICAL 

EXAMPLES AND THEIR SOLUTIONS 



DYNAMOS AND DYNAMO DESIGN 

DIRECT-CURRENT MOTORS 

ALTERNATING CURRENTS 

ALTERNATORS 

ALTERNATING-CURRENT APPARATUS 



SCRANTON: 

INTERNATIONAL TEXTBOOK COMPANY 

12B 



Digitized by VjOOQIC 



Copjrriflrfat. 1905. by Intbrnational Tbztbook Compamt. 



Entered at Stationers' Hall. London. 



Dynamos and Dynamo Desisrn: Copyriarht. 1906, by International Textbook 

Company. Entered at Stationers' Hall. London. 
Direct-Current Motors: Copyrisrht. 1906, by International Textbook Company. 

Entered at Stationers' Hall. London. 
Aitematinsr Currents: Copyrisrht, 1905, by International Textbook Company. 

Entered at Stationers' Hall. London. 
Alternators: Copyrisrht, 1905. by International Textbook Company. Entered 

at Stationers' Hall. London. 
Altematlnsr-Current Apparatus: Copyrisrht. 1905. by International Textbook 

Company. Entered at Stationers' Hall. London. 



All rifirhts reserved. 



Printed in the United States. 



//12b 

BURR PRINTING HOUSE, 
FRANKFORT AND JACOB STREETS. 

NEW YORK. 219 



Digitized by VjOOQIC 



104377 

MAR 3 1907 

SB 



PREFACE 



The International Library of Technology is the outgrowth 
of a large and increasing demand that has arisen for the 
Reference Libraries of the International Correspondence 
Schools on the part of those who are not students of the 
Schools. As the volumes composing this Library are all 
printed from the same plates used in printing the Reference 
Libraries above mentioned, a few words are necessary 
regarding the scope and purpose of the instruction imparted 
to the students of — and the class of students taught by— 
these Schools, in order to afford a clear understanding of 
their salient and unique features. 

The only requirement for admission to any of the courses 
offered by the International Correspondence Schools, is that 
the applicant shall be able to read the English language and 
to write it sufficiently well to make his written answers to the 
questions asked him intelligible. Each course is complete in 
itself, and no textbooks are required other than those pre- 
pared by the Schools for the particular course selected. The 
students themselves are from every class, trade, and profession 
and from every country ; they are, almost without exception, 
busily engaged in some vocation, and can spare but little 
time for study, and that usually outside of their regular 
w'orking hours. The information desired is such as can be 
immediately applied in practice, so that the student may be 
enabled to exchange his present vocation for a more con- 
genial one, or to rise to a higher level in the one he now 
pursues. Furthermore, he wishes to obtain a good working 
knowledge of the subjects treated in the shortest time and 
in the most direct manner possible. 

iii 



Digitized by VjOOQIC 



iv PREFACE 

In meeting these requirements, we have produced a set of 
books that in many respects, and particularly in the general 
plan followed, are absolutely unique. In the majority of 
subjects treated the knowledge of mathematics required is 
limited to the simplest principles of arithmetic and mensu- 
ration, and in no case is any greater knowledge of mathe- 
matics needed than the simplest elementary principles of 
algebra, geometry, and trigonometry, with a thorough, 
practical acquaintance with the use of the logarithmic table. 
To effect this result, derivations of rules and formulas are 
omitted, but thorough and complete instructions are given 
regarding how, when, and under what circumstances any 
particular rule, formula, or process should be applied ; and 
whenever possible one or more examples, such as would be 
likely to arise in actual practice — together with their solu- 
tions — are given to illustrate and explain its application. 

In preparing these textbooks, it has been our constant 
endeavor to view the matter from the student's standpoint, 
and to try and anticipate everything that would cause him 
trouble. The utmost pains have been taken to avoid and 
correct any and all ambiguous expressions — both those due 
to faulty rhetoric and those due to insufficiency of statement 
or explanation. As the best way to make a statement, 
explanation, or description clear is to. give a picture or a 
diagram in connection with it, illustrations have been used 
almost without limit. The illustrations have in all cases 
been adapted to the requirements of the text, and projec- 
tions and sections or outline, partially shaded, or full-shaded 
perspectives have been used, according to which will best 
produce the desired results. Half-tones have been used 
rather sparingly, except in those cases where the general 
effect is desired rather than the actual details. 

It is obvious that books prepared along the lines* men- 
tioned must not only be clear and concise beyond anything 
heretofore attempted, but they must also possess unequaled 
value for reference purposes. They not only give the maxi- 
mum of information in a minimum space, but this infor- 
mation is so ingeniously arranged and correlated, and the 



Digitized by VjOOQ IC 



PREFACE V 

indexes are so full and complete, that it can at once be 
made available to the reader. The numerous examples and 
explanatory remarks, together with the absence of long 
demonstrations and abstruse mathematical calculations, are 
of great assistance in helping one to select the proper for- 
mula, method, or process and in teaching him how and when 
it should be used. 

This volume contains an exceptionally clear and complete 
treatment of the design of direct-current dynamos and 
motors, together with a detailed discussion of the theory of 
alternating currents and descriptions of modern alternating- 
current machinery. In presenting the subject of design, a 
full discussion of the parts of the machines is first taken up, 
followed by complete demonstrations of design problems. 
The subject of armature windings receives special attention. 
Numerous winding diagrams are provided, indicating clearly 
the relative positions of the coils, pole pieces, and commu- 
tator bars. The text accords with the best modern practice. 
The theory of the action of motors, their connections and 
the many systems of speed control are set forth in a clear 
and comprehensive manner. The importance of a knowl- 
edge of alternating currents in the electrical-engineering 
profession . is steadily growing. Illustrations and the 
graphical methods of treatment have been freely used. 

The method of numbering the pages, cuts, articles, etc. 
is such that each subject or part, when the subject is divided 
into two or more parts, is complete in itself; hence, in order 
to make the index intelligible, it was necessary to give each 
subject or part a number. This number is placed at the 
top of each page, on the headline, opposite the page number; 
and to distinguish it from the page number it is preceded by 
the printer's section mark (§). Consequently, a reference 
such as § 16, page 26, will be readily found by looking along 
the inside edges of the headlines until § 16 is found, and 
then through § 16 until page 26 is found. 

International Textbook Company 



Digitized by VjOOQIC 



Digitized by VjOOQIC 



CONTENTS 



Dynamos and Dynamo Design Section Page 

Theory of the Dynamo 12 1 

Action of the Armature 12 3 

General Features 12 15 

Armature-Core Losses and Toothed Arma- 
tures 12 18 

Closed-Coil Armature Windings 12 20 

Manner of Winding the Coils 12 21 

Methods of Connecting Up Coils to the 

Commutator 12 26 

Parallel Windings 12 26 

Series-Windings 12 33 

The Magnetic Circuit 12 42 

Density of Lines of Force 12 44 

Form of Magnetic Circuit 12 45 

Methods of Exciting the Field 12 54 

Series-Winding 12 58 

Shunt Winding 12 62 

Compound Winding 12 66 

Building Up the Field 12 69 

Diagrams of Closed-Coil Windings ... 13 1 

Ring Windings . . , 13 1 

Drum Windings 13 4 

Parallel Windings 13 4 

Series-Windings 13 9 

Open-Coil Armature Windings 13 15 

Unipolar Dynamos 13 24 

Calculation of E. M. F. and Power ... 13 27 
Limiting Output of Constant -Potential 

Dynamos . . . . V 13 34 

iii 



Digitized by VjOOQIC 



Iv CONTENTS 

Dynamos and Dynamo Design — Continued Section Page 

Heating of Armature 13 35 

Sparking and Commutation 13 36 

Armature Reaction . 13 42 

Construction of the Armature 13 50 

Construction of Core and Spider .... 13 50 

Methods of Applying Windings 13 60 

Shafts 13 62 

Bearings 13 64 

Commutators 13 64 

Armature Losses and Heating 13 70 

Brown & Sharpe Gauge for Magnet Wire 13 72 

Design of the Field Magnet 13 77 

Magnetic Densities in Various Parts ... 13 79 
General Features Relating to Magnet 

Frames 13 81 

Determination of Ampere-Turns on Field 13 B3 

Field Windings 13 84 

Design of a 100-Kilowatt Dynamo .... 14 1 

Electrical Efficiencies of Dynamos .... 14 2 

Conditions Governing Preliminary As- . 

sumptions 14 3 

Heating Calculations 14 17 

Winding for 250 Volts 14 19 

Winding for 125 Volts 14 20 

Design of Commutator 14 21 

The Magnetic Circuit 14 23 

Computation of Field Windings 14 29 

Effects of Armature Reaction 14 33 

Calculation of Field Winding for 115-125 

Volts 14 37 

The Mechanical Design 14 42 

Summary of Dimensions 14 42 

Design of Armature and Commutator . . 14 43 
Construction of Field Frame and Field 

Coils 14 51 

Brush Holders and Rocker 14 58 

Bedplate and Bearings 14 62 



Digitized by VjOOQIC 



CONTENTS V 

Dynamos and Dynamo Design — Continued Section Page 

Connections 14 64 

Efficiency 14 67 

250- Volt and 600- Volt Generators .... 14 70 

Testing 14 72 

Direct-Current Motors 

Principles of Operation 15 1 

Dynamos and Motors Compared .... 15 1 

Action of Motor 15 2 

Counter E. M. F. of Motor 15 3 

Motor Efficiency 15 8 

Commercial Efficiency of Motors .... 15 10 

Torque 15 11 

Armature Reaction 15 17 

Classes of Motors 15 19 

Shunt Motors 15 20 

Speed Regulation of Shunt Motors ... 15 21 

Series Motors 15 24 

Series Motor on Constant-Potential Circuit 15 24 

Speed Regulation on Series Motor ... 15 28 

Series Motor on Constant-Current Circuit 15 29 

Compound- Wound Motors 15 30 

Differentially Wound Motors 15 30 

Accumulatively Wound Motors 15 31 

Dynamo and Motor Rotation 15 32 

Auxiliary Apparatus 15 35 

Starting Rheostats 15 35 

Shunt-Motor Connections 15 37 

Reversing Direction of Rotation ... 15 46 

Series-Motor Connections 15 50 

Automatic Starting Rheostats 15 56 

Multivoltage Speed Control 15 59 

Teaser System of Control 15 65 

Control by Variation of Field Reluctance 15 67 

Design of Direct-Current Motors .... 15 68 

Determination of Output .15 69 

Design of 10-Horsepower Shunt Motor . 15 71 



Digitized by VjOOQIC 



vi CONTENTS 

Direct-Current Motors — Continued Section Page 

Design of 10-Horsepower Series Motor . 15 72 

Mechanical Design 15 73 

Stationary Motors 15 73 

Care and Operation of Dynamos and Motors 15 75 

Brushes 15 76 

The Commutator 15 78 

The Armature 15 80 

Field-Coil Defects 15 82 

Reasons for Dynamo Failing to Generate 15 84 

Failure of Motor to Start 15 88 

Sparking 15 89 

Testing for Faults 15 92 

Alternating Currents 

E. M. F. Wave Forms 16 1 

Cycle, Frequency, Alternation, Period . . 16 4 

Sine Curves 16 6 

Properties of Sine Curves 16 9 

Addition of Sine Curves 16 9 

Two-Phase and Three-Phase Systems . . 16 18 
Composition and Resolution of Currents 

and E. M. F.'s 16 20 

Maximum, Average, and Effective Valves 

of Sine Waves 16 22 

Relations between Values 16 25 

Self-Induction and Capacity 16 28 

Circuits Containing Resistance Only ... 16 30 
Circuits Containing Self-induction Only . 16 31 
Circuits Containing Resistance and Self- 
Induction 16 38 

Angle of Lag 16 42 

Circuits Containing Capacity Only .... 16 43 

Circuits Containing Resistance andCapacity 17 1 
Circuits Containing Self-induction and 

Capacity 17 6 

Circuits Containing Resistance, Self-induc- 
tion and Capacity 17 18 



Digitized by VjOOQIC 



CONTENTS vii 

Alternating Currents — Continued Section Page 
Calculation of Power Expended in Alter- 
nating-Current Circuits 17 18 

Power Factor of a Circuit 17 25 

Wattless and Power Components .... 17 26 

Transmission Lines 17 28 

Alternating - Current Measuring Instru- 
ments 17 34 

Classes of Instruments 17 35 

Hot-Wire Ammeters and Voltmeters . . 17 35 

Plunger and Magnetic- Vane Instruments .17 39 

Induction Instruments 17 41 

Electrodynamometers 17 44 

Wattmeters 17 48 

Electrostatic Voltmeters .17 64 

Alternators 

Sine-Phase Alternators 18 1 

Construction of Alternators 18 5 

Alternators 18 12 

Calculation of E. M. F. Generated by 

Alternators 18 14 

Field Excitation of Alternators 18 20 

Revolving-Field and Inductor Alternators 18 25 

Polyphase Alternators 18 32 

Two-Phase Alternators 18 32 

Three-Phase Alternators 18 39 

Star and Delta Connections 18 42 

Relation Between Current, E. M. F., and 

Output 18 46 

Monocyclic System 18 51 

Alternators With Closed-Circuit Armature 

Windings . ^. 18 52 

Alternating-Current Apparatus 

Transformers 19 1 

Theory of the Transformer 19 3 

Action of the Ideal Transformer .... 19 8 



Digitized by VjOOQIC 



viii CONTENTS 

Alternating-Current Apparatus — Continued 

Section Page 
Effect of Resistance of Primary and Sec- 
ondary Coils 19 11 

Effect of Magnetic Leakage 19 11 

Effect of Core Losses 19 12 

Construction of Transformers 19 14 

Examples of Transformers 19 15 

Alternating-Current Motors , . 19 23 

Synchronous Motors 19 23 

Induction Motors 19 28 

Methods of Starting Induction Motors . . 19 41 

Field Connections 19 49 

Single-Phase Induction Motors 19 52 

Series Motor on Alternating Current . . 19 57 

Shunt Motor on Alternating Current ... 19 57 

Repulsion Motor 19 58 

Wagner Single-Phase Induction Motor . . 19 59 

Rotary Converters 19 64 

Single-Phase Converters 19 64 

Two-Phase Converters 19 65 

Three-Phase Converters 19 66 

Multipolar Rotary Converters 19 69 

Operation of Rotary Converters 19 71 

Double-Current Generators 19 77 



Digitized by VjOOQIC 



DYNAMOS AND DYNAMO DESIGN 

(PART 1) 



THEORY OF THE DYNAMO 



IKTRODFCTION 

!• Principles of Construction. — It has been shown 
that when an electrical conductor cuts across a magnetic 
field an E. M. F. is induced in the conductor. If the cir- 
cuit is completed between the ends of the moving conductoi^ 
a current will be established as shown in Fig. 1, which cur- 
rent will, in turn, react on the 
magnetic field and exert a 
force tending to stop the 
motion of the conductor — the 
arrow a shows the direction 
of motion while b shows the 
direction of the reacting force. 
If we continue to move the 
conductor against this force, '°' * 

the current will continue to flow, but work must be done to 
move the conductor against the opposing force ; thus it might 
be said that we have converted dynamic energy into elec- 
trical energy, for electrical energy has been obtained at the 
expense of mechanical effort. A machine for generating 
electrical energy that operates on this principle is called 
a dynamo-electric machine or an electric generator. 
These terms are often abbreviated in dynamo and 
Srenerator. 

For notice of copyright, see page immediately following the title page. 
44—3 




Digitized by VjOOQ IC 



2 DYNAMOS AND DYNAMO DESIGN §12 

2. Essential Parts of a Dynamo. — The simplest of all 
mechanical motions is that of rotation, and dynamos always 
use this principle for sweeping the conductors through the 
magnetic field. There are essentially two parts to a dynamo: 
first, the field magnet, wherein is produced the necessary 
magnetism, and second, the armature, on or near whose 
surface the working conductors are arranged. These two 
parts are rotated relatively to each other, it being imma- 
terial, except for convenience, which is stationary and which 
is rotated. 

3. It is seldom that a single conductor can be made to 
generate a desired voltage, so there are usually on an arma- 
ture a number of conductors that are connected up in series 
and in parallel, in the same way as electric batteries, until 
the required voltage and current-carrying capacity are 
obtained. The subject treating of the methods of intercon- 
necting armature conductors is called arviature windings 
and will be treated in detail later. 

4. Classes of Dynamos. — Dynamos are divided into 
two classes according to the character of the current they 
generate, namely, direct -current dynamos^ abbreviated D. C, 
and alternating-current dynamos^ abbreviated A. C. Direct- 
current machines deliver currents that are continuous in 
direction, though perhaps varying in amount as required, 
while alternating-current machines deliver currents that 
periodically reverse in direction many times per second. 
Alternating currents and the dynamos for producing them 
will be treated separately, the present section being confined 
to direct-current dynamos. 

5. The direct-current dynamo necessarily has a stationary 
field magnet and a revolving armature. The reason for this 
is that the brushes that collect the current generated in the 
armature must keep a fixed position with respect to the poles 
of the field magnet, and since these brushes require occa- 
sional adjustment it is best that they remain stationary, 
hence the field-magnet poles must also be stationary. 



Digitized by VjOOQIC 



§12 DYNAMOS AND DYNAMO DESIGN 3 

6. The field magnet may be either a permanent magnet 
or an electromagnet; but except for very small dynamos, 
electromagnets are necessary because they are more power- 
ful and because it is practically impossible to construct and 
harden large masses of steel of a kind suitable for permanent 
magnets. Where the field is produced by electromagnets, 
the exciting coils are usually supplied with current generated 
by the dynamo's own armature; such a machine is termed a 
self-excited dynamo. Where an outside source of current 
is resorted to for excitation the machine is called a sepa- 
rately-excited dynamo. 

Quite a number of shapes and styles of field magnets have 
been successfully used commercially, and will be discussed 
later. Like other magnets, there must be a complete mag- 
netic circuit, but for the purpose of the following discussion 
only the armature and pole pieces will be considered, the 
remainder of the magnetic circuit being for the present 
omitted. 

ACnON OF THE ARMATURE 

7. In Fig. 2 is shown an end view of the pole pieces 
N and 5 of a field magnet, 'while between them is shown an 




PiO. 8 



armatnre core A of soft wrought iron mounted on a suit- 
able shaft and capable of being rotated. This wrought-iron 



Digitized by VjOOQIC 



4 DYNAMOS AND DYNAMO DESIGN §13 

armature core being of high magnetic permeability serves 
to convey or conduct the magnetic induction or magnetic 
flux from pole to pole, greatly reducing the magnetic reluc- 
tance of the path. The space between the core and the 
poles is called the air grap. This space is often occupied 
with insulation, copper wires, etc., but since these sub- 
stances have practically the same permeability as air, the 
complete region from pole to armature core is referred to as 
the air gap, regardless of whatever may intervene, so long 
as it is not iron. 

It will be seen that a conductor c attached to this arma- 
ture core and revolved with it in the direction indicated, will 
cut across all the magnetic induction either entering or 
leaving the core, and, consequently, will have an E. M. F. 
induced in it. Applying the rule for determining the direc- 
tion of the induced E. M. F., it will be found that when the 
conductor is under a north pole, as at ^, the E. M. F. is 
upwards from the paper, while when it is under the south 
pole it is downwards. If the direction of rotation were 
reversed, the E. M. F. would be upwards under the south 
pole and downwards under the north. In all the diagrams 
relating to dynamos or motors, an up-flowing current, that 
is, a current flowing up through the plane of the paper, will 
be represented by a dot in the center of the wire as at a\ 
down-flowing currents will be represented as at d, the wire 
being filled in black. 

It will be noticed that the wrought-iron core need not be 
rotated in order that the conductor may be caused to gener- 
ate an E. M. F., but no serious difficulty is encountered in 
so doing, and it is very convenient to support the conduc- 
tors by attaching them firmly to the core and rotating the 
whole. 

8. In the conductors. Fig. 2, there is induced an E. M. F. 
that alternates in direction every half revolution ; such an 
E. M. F. is called an alternating: E. M. F., and would 
induce an alternating current were the circuit completed 
between the ends of the conductor. 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 6 

Suppose another conductor dwere also attached to the core 
diametrically opposite to ^. It also would have E. M. F.'s 
induced in it in the same manner as c; but, being on oppo- 
site sides when the E. M. F. of one was upwards, that of 
the other conductor would be downwards, and vice versa. 
These two conductors could then be connected together at 



Pio. 8 

one end of the armature, as in Fig. 3, without having 
their E. M. F.'s interfere or at any time oppose one another 
so that at the ends of the loop, or turn, the sum of the two 
induced E. M. F.'s would be impressed. 

9. Elementary Alteruatingr-Current Generator. — In 

Fig. 3 the ends of the loop are shown as terminating in two 
metal rings insulated from each other and from the shaft. 
Upon these collector ringrs ^, A, as they are called, rub 
two metal brushes e, f that serve to collect the current 
generated in the armature wires and make it available for 
the outside circuit R. The E. M. F. impressed on the cir- 
cuit R is still an alternating E. M. F., since the brushes 
make permanent, though sliding, connection with the 
loop ^, d\ hence, the current that flows through R will be an 
alternating current. In fact. Fig. 3 represents an elemen- 
tary alternating-current generator. 



Digitized by VjOOQ IC 



6 



DYNAMOS AND DYNAMO DESIGN 



gl2 



10. In Fig. 4, a curve is plotted that shows the relation 
between the E. M. F. and the time required for a revolution, 
using the volts between the brushes ^, f as ordinates 
and the time in seconds as abscissas. Starting with the 




loop in a position midway between the poles, there is no 
E. M. F. developed, since no cutting takes place until 
the face conductors pass beneath the pole faces. As the coil 
passes under the poles, the E. M. F. rises, reaching a maxi- 
mum and remaining there until the loop again passes out 
between the poles, when the E. M. F. falls to zero and later 
rises in the reverse direction, etc., as shown in Fig. 4. One 
positive wave, represented above the axis, and one negative 
wave of E. M. F., represented below the axis, are developed 
every revolution. It should be noted that the shape of the 
curve in Fig. 4 depends on the distribution of the magnetic 
field around the armature. 

11. Suppose the collector rings ^, //, Fig. 3, were replaced 
by a single ring split into two semicircles, the halves being con- 
nected, each to an end of the loop, as in Fig. 5, but insulated 
from the shaft and from each other. A current of entirely 
different character would now be collected by the brushes e^f^ 
for as the loop revolves so also does the split ring, and the 
brushes and split ring are so arranged that, at the instant 
when the E. M. F. induced in the loop reverses, the brushes 
reverse their connections to the loop by sliding from 



Digitized by VjOOQ IC 



§12 



DYNAMOS AND DYNAMO DESIGN 



one-half of the split ring to the other. In this way the 
E. M. F. at the brushes will not reverse at each half revolu- 
tion, although the E. M. F. induced in the loop is alternating 




Pxo. 6 

as before. In Fig. 6 is shown the corresponding E. M. P. time 
curve. Comparing it with the curve in Fig. 4, it will be 
noted that they are the same, except every alternate wave 




PIO. 6 



has been reversed in direction by the rotation of the split 
ring and the consequent change of connection with the 
outside circuit R, 



Digitized by VjOOQIC 



8 DYNAMOS AND DYNAMO DESIGN §13 

12. An E. M. F. like that plotted in Fig. 6, since it does 
not alternate in direction, is called a direct E. M. F., but 
on account of its variable character it is usually termed a 
pulsating: E. M. F. The variations are due entirely to the 
fact that there is but a single loop on the armature of 
Fig. 6, which, from the nature of the case, cannot always be 
generating. The E. M. F. may be made quite uniform by 
using several loops instead of one, connecting them up after 



Pig. 7 

the manner of Fig. 7. This shows two loops connected to a 
ring split into four segments. The armature core and shaft 
are omitted in order to make the windings more distinct. 
The loop mm' terminates in the opposite segments ^, a^ 
while the loop n n' terminates in the segments ^, b\ 

13. The action of the two-loop armature in smoothing 
out the pulsations is as follows: When the face conductors 



Digitized by VjOOQ IC 



§ 12 DYNAMOS AND DYNAMO DESIGN 9 

of one loop, which has been active, nears the edges of the 
pole pieces where its E. M. F. will fall off, the segments to 
which it is connected slide from under the brushes and the 
loop is cut out of circuit ; at the same time the other loop, 
which is just approaching the most active position, is cut 
into the circuit, maintaining the E. M. F. during the inac- 
tivity of the first coil and carrying whatever current the 
dynamo is generating until it, in turn, is cut out, and so on. 
The loops are only in the circuit while they are in active 
positions, so that the E. M. F. cannot fall much in the out- 
side circuit, and the pulsations are therefore greatly reduced. 
In Fig. 7, +5 and —B are the brushes and Re, the external 
circuit to which the armature supplies current, as indicated 
by the ammeter A, M. 

14, Cominutator. — The split ring is called the conunu- 
tator, and it is the essential and distinctive feature of the 
direct-current dynamo, for it is by its use that the alterna- 
ting E. M. F.'s developed in the armature windings are 
rectified, or commuted into direct E. M. F.'s. In most 
modern generators the commutators have a great many 
segments. 

If a single loop or turn of wire, as shown in Figs. 5 or 7, 
does not develop the required E. M. F., more turns may be 
added, and each additional turn adds two active conductors 
in series with the others. If all the turns of a coil are 
wound approximately in the same plane, the E. M. F. *s of 
all face conductors will rise and fall together, so that the 
action of what has now become a coil is identical with that 
described for a single loop, except that the E. M. F. 's are 
increased in proportion to the number of turns in the coil. 
Diagrams of windings, then, like Figs. 5 or 7, may be consid- 
ered as having coils of many turns instead of one, with the 
ends of the coils terminating in segments as shown for the 
single turn, and in many of the diagrams to be given later 
it will be understood that where but a single turn per coil 
is shown for simplicity, many turns per coil will have iden- 
tically the same action. 



Digitized by VjOOQ IC 



10 



DYNAMOS AND DYNAMO DESIGN §12 



16. Drum and King: Armatures. — Windings made 

after the manner of Figs. 5 
or 7 are called drum wind- 
ing^, on account of the arma- 
ture core being of a drum 
form, and an armature thus 
wound is called a drum-iivound 
armature, or a drum arma- 
ture. Another type of wind- 
ings not nearly so generally 
used is wound on a core made 
into the form of a ring with the 
windings threaded through the 
ring, as shown in Fig. 8. It 
will be seen that where more 
than one turn is put on, as in 
Fig. 8, that the E. M. F.'s de- 
veloped will be in series exactly 
as are the E. M. F. *s on the two sides of the loop in Fig. 3. 

16. Inductors, or Face Conductors. — Fig. 9 repre- 
sents the magnetic flux around a ring-wound armature. It 




Pio. 8 




PIO. 9 



will be noticed that but a small field is found in the middle 
of the ring, so that, as in the case of the drum winding, the 



Digitized by VjOOQIC 



§12 DYNAMOS AND DYNAMO DESIGN 11 

part of the conductor that generates the E. M. F. is the part 
that is on the outside of the ring and sweeps under the 
poles. This part of a conductor is usually termed an 
Inductor, or a face conductor. It is evident that a face 
conductor on a ring core will develop the same E. M. F. as 
one on a drum core, where the total induction or flux from 
one pole to the other is the same and where the speed is the 
same. Thus, the coil of two turns in Fig. 8 is exactly 
equivalent to one loop or turn in Fig. 7, so far as the gen- 
eration of an E. M. F. is concerned, since in each case there 
are two face conductors in series. 

Fig. 10 shows a ring winding exactly equivalent, elec- 
trically, to the drum winding of Fig. 7, except that it will 



PlO. 10 

develop four times the voltage of Fig. 7, at the same speed, 
because it has four times as many face conductors in series 
between opposite commutator segments. 

17, Opon-Coil and Closed-Coil Windingrs. — The wind- 
ings shown in Figs. 5, 7, and 10 are known as open-coil 
winding^, from the manner in which they are connected 
to the commutator. Another type, the closed coil, which 
is far more used, is shown in diagram in Fig. 11. This 
shows a ring wound continuously by a conductor whose 



Digitized by VjOOQ IC 



12 



DYNAMOS AND DYNAMO DESIGN 



§12 



ends are joined, thus forming a closed winding, whence the 
name. As the armature revolves, conductor a has an 
E. M. F. induced in it which is added to that of b^ c, d, and 
g, because they are all connected in series, and the difference 
of potential between the points x and y is the sum of the 
E. M. F/s developed in all the conductors under the iVpole. 
In the same way, between o and / is the sum of the 
E. M. F.'s developed under the S pole, and since these poles 
are of the same strength and size, and since the conductors 
are evenly spaced, there are as many conductors from x to y 
as from otop and the E. M. F. between x and^ is equal to 




Pio. u 

that between o and p. It will be noticed that these two 
induced E. M. F.'s are opposed to each other in the windings 
and, being exactly equal, no current can be^produced in the 
armature itself ; but if a pair of brushes e, f are arranged 
so that they will rub on the conductors as they pass what 
is termed the neutral region between the poles, e will be 
at the same potential as / and j/, and f at the potential of 
and ;r, and if a circuit R is completed between these 
brushes, a current will be set up. Notice that, within the 
armature, the current divides as shown by the arrows ; it 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 13 

would be said of this armature winding that there are two 
circuits, or paths, from positive to negative brushes. As 
the armature revolves, the conductors successively come in 
contact with the brushes, but the number of conductors 
under each pole (on which the E. M. F. between the brushes 
depends, assuming the total flux and the speed to remain the 
same) remains always about the same, so that the E. M. F. 
will remain constant and the action will be continuous. 

It is not always convenient or possible to have the brushes 
rub on the conductors themselves, and the winding is there- 
fore, in nearly every 

case, connected to a ** 

commutator as 
shown in Fig. 12. 

18, Diagrams 
Figs. 10 and 12 are 
typical of ring-wound 
open-coil and closed- 
coil windings, respect- 
ively, and should be 
carefully compared. 
Notice that the ends 

of a coil in Fig. 10 " 

^ PIO. 12 

terminate in s e g - 

ments on opposite sides of the commutator, while in Fig. 12 
the ends of a coil connect to adjacent commutator seg- 
ments. Also, in Fig. 10, the brushes are in connection 
with a coil when it is under the poles, while in Fig. 12 the 
brushes are only in contact with a coil when it is in the 
neutral region. 

19, Conmiutatlon. — Consider the coil b in Fig. 12. It 
will be noticed that it is short-circuited, at the instant 
shown, by the brush e touching both segments in which 
the coil terminates. The coil a to the right of it has the 
current flowing through it to the left, while that to the 
left c has the current flowing through it to the right. As 
the coils are continually passing from right to left as the 



Digitized by VjOOQ IC 



14 DYNAMOS AND DYNAMO DESIGN §12 

armature revolves, the current in each must be reversed as 
it passes the brush. In a closed-coil winding, then, as the 
segments in which a coil terminates pass a brush, the coil is 
short-circuited for a short interval, and, further, the direc- 
tion of the current in it is reversed. This action is termed 
commutation, and coil b is said to be under commutation 
at the instant shown. 

In Fig. 13 is shown a drum-wound closed-coil type of 
armature winding with 8 coils, each having two face con- 



B 

Pig. 18 

ductors. These coils have their ends terminating in adja- 
cent commutator segments; hence, it is a winding similar in 
action to Fig. 12. In fact. Fig. 13 is, electrically, exactly 
equivalent to Fig. 12, except that the drum type of coil is 
used in one and the ring in the other. They would generate 
exactly the same voltage if run at the same speed with the 
same magnetic induction from pole to armature, because 
both have the same number of face conductors per coil and 
the same number of coils. In the diagram, the connections 
across the rear end of the core between conductors are 
omitted in order not to confuse the figure. A conductor on 
one side does not connect to the one diametrically opposite 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 16 

it, but lacks one conductor of doing so, thus forming what 
is sometimes called a chord Tending. The turn connects 
across a chord instead of a diameter, thus making the turn 
slightly shorter and avoiding the armature shaft to better 
advantage. Drum armature windings will be taken up in 
detail in connection with the subject of armature windings. 

80, Armatures are said to be drum armatures when sup- 
plied with a drum type of winding and ring armatures when 
supplied with a ring type of winding, regardless of the shape 
of the armature core. Almost all modern machines use 
armature cores that are of the ring shape because these 
machines have armatures of large diameter and short length, 
but if the winding does not thread through the center of 
the ring it is a drum winding and the armature is called 
a drum armature. 

OENSBAXi FEATURBS 

21, Multipolar Field Magrnets. — Dynamos whose field 
magnets have but a single pair of poles are called bipolar 
dynamos, while those with 
more than one pairof poles are 
called multipolar. In mul- 
tipolar machines the poles 
always alternate north and 
south, as in Fig. 14, so that 
it is necessary to have as 
many north as south poles, 
consequently there is always 
an even number of poles. 

23. Direct-Driven and 
Belt-Driven Dynamos. 

Dynamos, like any other kind ^°- ^* 

of machinery, are driven by steam engines or waterwheels 
and may either be connected by a belt and pulley or may be 
direct connected to the prime mover. The former are 
termed belted dynamos and the latter, where the armature 
is mtended to be mounted on an extension of the shaft of a 




Digitized by VjOOQ IC 



16 DYNAMOS AND DYNAMO DESIGN §12 

steam engine, are termed direct-driven or engflne type. 
The former type usually run at higher speeds than the 
latter, but otherwise the two classes of machines are identical. 



o 



Fig. 15 shows a modern belted dynamo. A is the circular 
yoke of the field magnet, which is made in two pieces for 



Digitized by VjOOQIC 



la DYNAMOS AND DYNAMO DESIGN 17 



8 
I 



^ 44—3 



Digitized by VjOOQIC 



18 DYNAMOS AND DYNAMO DESIGN § 12 

convenience in building and repairing. The pole pieces pro- 
ject inwards from A and are surrounded by the magnetizing 
coils By termed the field colls. The armature H revolves 
within the poles, being supported by a shaft that also carries 
the pulley for driving. The armature windings may be 
seen terminating in the segments of the commutator C, on 
whose surface rub the brushes J/, which serve to collect 
the current. The devices for holding the brushes are 
called brusli holders, and these are supported on an insu- 
lated stud N securely fastened to the rocker-arm K. The 
brushes require occasional adjustment and the rocker-arm 
is capable of rotating through a small angle, the movement 
being controlled by a worm-gear, not shown, connected to 
the hand wheel J. 

The shaft is supported by three pedestals G^ which contain 
the bearings and their oiling devices. The bedplate F to 
which the pedestals are bolted is supported on rails/?, two of 
which are provided with screw devices for sliding the com- 
plete dynamo for the purpose of varying the tension on the 
driving belt. Fig. 16 shows a comparatively small direct- 
driven dynamo. The dynamo bedplate A is practically an 
extension of the engine bedplate. The armature AT is 
mounted on the engine shaft ; C, D are the terminals of the 
machine and W the hand wheel for regulating the brushes. 



ABMATUBE-CORE LOSSES AND TOOTHED ARMATURES 

23. Eddy-Current lioss. — Thus far armature cores have 
been considered as made of a solid mass of soft wrought iron, 
having the conductors forming the winding attached to the 
surface. Some of the early dynamos were thus constructed 
but they were found to be inefficient on account of currents 
being induced in the iron core itself as well as in the con- 
ductors. These currents circulating in the mass of iron 
caused the core to heat on account of the resistance offered 
to the currents. They are called eddy currents and the 
loss they entail is called the eddy-current loss. 



Digitized by VjOOQ IC 



§12 



DYNAMOS AND DYNAMO DESIGN 



19 




PIO, 17 



The E. M. F. producing these eddy currents is necessarily 
low, but if the core is a solid mass of metal the resistance 
offered is extremely small and 
the small E. M. F. will cause 
enormous currents to flow. 
The direction of these cur- 
rents near the pole surface is 
the same as those induced in 
any other conductor and the 
currents circulate as shown in 
Fig. 17, which shows half of 
an armature core only, with 
the direction of the eddy currents marked on the section. 
To reduce the eddy currents and thus prevent the losses 
they entail, armature cores are built up of a number of thin 

iron disks from .01 inch to 
.06 inch thick, as shown in 
Fig. 18, arranged parallel to 
the lines of force and perpen- 
dicular to the axis of rotation. 
These disks are insulated 
from one another, and the 
length of conductor parallel 
to the shaft being reduced to a small fraction of an inch 
reduces the magnetic flux cut to a very small value of what 
it would be were there several inches of solid conductor; 
hence, the rate of cutting and the E. M. F. tending to set 
up eddy currents is greatly reduced. Further, the resistance 
offered to the flow of these currents is so increased that the 
currents themselves are doubly reduced. 

This process of dividing the core into thin plane sections 
is called lamination, the separate sections forming the 
lamince. Lamination does not affect the magnetic qualities 
of the core, since all the sections are continuous in the 
direction of the lines of force. 

Building up the core of lightly insulated iron wire will 
also prevent eddy currents, but as in this case the iron of 
the core is not magnetically continuous, the reluctance of 




Pig. 18 



Digitized by VjOOQ IC 



20 DYNAMOS AND DYNAMO DESIGN §13 

the core as a whole is much greater than that of the iron 
of which it is composed. The laminated structure is there- 
fore used almost exclusively. 

24, Hysteresis Ijoss. — Revolving the armature core in 
the magnetic field entails still another loss of energy due to 
magrnetic hysteresis. Hysteresis is due to the continual 
change in the direction of the lines of force through the core 
as it rotates, amounting to one complete reversal in each 
half revolution in a bipolar machine. The amount of the 
hysteresis loss depends on the quality of the iron of which 
the core is made, the density of the lines of force in the core, 
the number of reversals of the magnetism per second, and the 
amount of the iron affected. 

25. The chief reluctance of the magnetic circuit of a 
well-designed dynamo is at the air gaps, and in order to 
reduce this to a minimum the conductors are usually laid in 
slots, or even in holes near the surface, so that the iron of 
the core approaches very nearly to that of the pole pieces. 
Armature cores where the windings are so made are termed 
slotted, or toothed, armatures and perforated arma- 
tures, respectively, while if surface windings are used they 
are called smooth-core armatures. One of the advan- 
tages to be gained by laying the windings in slots is that 
the conductors and their insulations are thus thoroughly 
protected from injury. 



CLOSED-COIL ARMATURE WINDINGS 

' 26. Closed-coil armature winding usually consist of 
a large number of coils connected up to a commutator with 
many segments. These coils may consist of one or of many 
turns, according to the requirements and may be either drum 
or ring wound. It should be understood that the manner of 
winding the coils is entirely separate and distinct from the 
manner of connecting them to the commutator. Windings 
differ very much in their properties according to the method 
of interconnecting the coils with the commutator. 



Digitized by VjOOQ IC 



12 



DYNAMOS AND DYNAMO DESIGN 



21 



27. The coil is the natural unit of the winding and for 
the present purpose may be defined thus: If we start at any 
commutator segment and follow a conductor connected 
thereto around through its convolutions on the armature 
until we arrive at another segment, we will have traversed 
a coil. 



MANNER OF WINDING THE COLLS 

28, Pitch, or Spread, of Coils. — As has already been 
stated, the drum type of coil is by far the most used. In 
Fig. 19 is shown a drum- 
wound coil in a multipolar 
field. The sides of the coil a 
and b should be separated 
by about the same angle c 
as the angle between adja- 
cent poles d in order that 
the conductors at a may 
pass under and out from the 
poles simultaneously with 
the conductors at b. The 
angle c is often referred to 
as the angular pitch, or 
spread, of the coils, and 
the angle d^ which is the 
angle between corresponding points on adjacent poles, as the 
angular pitch of the i>oles. The angular pole pitch in a 
bipolar machine is 180°, in a four-pole machine 90°, etc. 
The spread of the coils may differ a little from the angular 
pole pitch without affecting the action of the machine, but 
if the difference is large the coil will not commutate prop- 
erly. This is because the coil is short-circuited by a brush 
during commutation and this should take place only when 
the coil is not developing any E. M. F., that is to say when 
both sides are in the neutral region between the poles. Now 
it will be seen on consideration that with poles of a given 
size, the angle through which the armature and coil can turn 




Pio. 10 



Digitized by VjOOQ IC 



22 



DYNAMOS AND DYNAMO DESIGN § 12 



while both sides remain in the neutral region is greatest 
when the spread of the coils is equal to the pole pitch. 

On most armatures there are a large number of coils that 
overlap one another considerably. On account of sparking 
at the commutator, it is very desirable that all armature coils 
should be alike; this is especially desirable with multipolar 
armatures. If a coil were wound and then another over it, 
and so on, the first coil, being underneath, would be much 
shorter than the last one at the top. This would cause 
unequal division of the currents through the paths, or cir- 
cuits, in the armature and unequal heating due to different 
resistances on the various paths. In order to avoid this 
trouble, multipolar drum windings are usually made up of 
coils formed on a frame into such shapes as will nest together 
and are afterwards assembled on the core. 




FIO. 90 



29, Shape of Colls. — There are two chief methods of 
making these coils, both of which are practically alike, a 




Pig. 21 



typical coil of each is shown in Figs. 20 and 21. The 



Digitized by VjOOQIC 



§12 DYNAMOS AND DYNAMO DESIGN 23 

parts a a and b b are composed of the face conductors and lie 

in the slots, the parts d^ e are 
termed the end connections and 
project beyond the armature 
core. The terminals /, /, termed 
the leads (pronounced leeds)^ 
serve to connect the coil to the 
commutator. It will be seen 
that the part a a lies in a differ- 
ent plane than the part bb^ the 
turn at c serving to connect the 
two. The part a a therefore lies 
in the bottom half of some slot 
and the part ^ ^ in the top half 
of some other slot. The coils 
lie in two layers and constitute 
what is sometimes called a two- 
layer winding. This is done 
in order to keep the end connec- 
. tions from interfering, for all 
£ the end connections e, from the 
left-hand side of each coil, ex- 
tending toward the right, are 
beneath the end connections, 
from the right-hand side d^ ex- 
tending toward the left. This 
will be more readily understood 
from Figs. 22 and 23, which 
show a few coils of each type on 
the armature core. In Fig. 23 
the right-hand sides of the two 
coils have not been forced down 
into position in the slot. 

In Fig. 22, it will be seen that 
the end connections beyond the 
slots are supported by flanges 
that give the completed armature 
a cylindrical appearance, from 



Digitized by VjOOQ IC 



24 DYNAMOS AND DYNAMO DESIGN §12 

which it is called a barrel-AVOund, or cyllndrlcal-'wound, 



armature. In Fig. 23, the ends of the coils bend down 

over the end plattes 
of the core instead 
of projecting straight 
out. In Fig. 24 is 
shown an end view 
of a complete arma- 
ture wound with coils 
similar to Fig. 21; 
this type of winding 
is appropriately 
called a spiral, or 
involute, . winding. 
The various turns ma- 
king up the individ- 
ual armature coils are 
Fio. 94 held in place by tape. 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 25 

30. Ring-wound coils are almost invariably made like 
Fig. 12, except that the face conductors are usually wound 
in slots in the armature core. Sometimes ring-wound coils 




PIO. 96 

are connected two in series, as shown in Fig. 25. By the 
definition of a coil previously given, this would be but a 
single coil, although composed in reality of two coils a and b 
connected in series. 

The only advantage this type of coil has over the plain 
ring-wound coil is that less commutator segments are 
required than if the same number of parts of coils were 
connected up as plain ring coils. This is obviously so, since 
there are only half the number of coils that there are parts 
of coils. This winding is, in action, exactly the same as a 
drum winding, for it has face conductors under each of two 
adjacent poles and what was stated about the proper spread 
of drum-wound coils also applies to this winding. Notice 
that the direction of winding the two parts is different so 
that the E. M. F. developed under opposite poles will be 
additive in the coil. The winding of Fig. 10 is of this 
character. 

The chief advantage of the drum type of coil is that 
it can be wound on a form, insulated, and then assem- 
bled on the core, while the ring type must be wound in 
place and cannot be so perfectly insulated at the same 
expense. 



Digitized by VjOOQIC 



26 



DYNAMOS AND DYNAMO DESIGN 



§12 



METHODS OF CONNECTING UP COLLS TO THE 
COMMUTATOR 



31. Single Parallel Winding. — The simplest of the 
methods of connecting up the coils to the commutator is 
that shown in Figs. 12 and 13, in which the ends of each 
coil are connected to adjacent commutator segments. This 
winding is called a i>arallel or a multicircuit winding. 

32. In Fig. 26 is shown such a winding arranged for 
six poles. The coils themselves are not shown but only 



If 



Fig. 26 



indicated by the loops from segment to segment. It is imma- 
terial whether these coils are of the drum or ring type so 
long as they conform to the requirements already explained 
for armature coils. In Fig. 26 the brushes are shown 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 27 

rubbing on the inside of the commutator, in order to avoid 
confusing the diagram. * 

Note that to each segment there are connected two leads 
from two different coils. It might be said that to each seg- 
ment were connected the end of one coil and the beginning 
of the next, or the right-hand end of one coil and the left- 
hand end of the next. Now there are but two ends to each 
coil, and since two ends, or leads, are connected to each com- 
mutator segment, it follows that there are just as many 
segments as coils. This is true for all closed-coil windings 
of whatever kind. 

33. Position of Brushes. — 'Suppose that in Fig. 26 the 
brush a receives 100 amperes from the negative terminal. 
The current will enter both segments P and (?, dividing 
equally, if the winding is symmetrical, so that 50 amperes 
will flow through the coils O N-N M-M L, etc., and 
50 amperes will flow through the coils P Q-QR-RS^ etc. 
As the armature rotates, the coil PO will pass to the right 
and 50 amperes current will then flow in it in the direction 
PO. A moment before the position in the figure was reached, 
the coil was to the left and had 50 amperes in it flowing in the 
direction OP. Thus a current of 50 amperes is reversed 
in every coil as the segments to which it connects pass 
the brush ^, and further it is because the brush delivers 
100 amperes to the windings that the current of 50 amperes 
is reversed. This is evidently true, for if both sets of coils 
to the right and to the left are to have currents in them flow- 
ing away from ^, the brush a must supply the sum of these 
currents. 

34. The E. M. F. generated by a coil reverses when that 
coil passes from under one pole to one of opposite polarity 
and the current in the coil should reverse at the same time; 
hence, in all closed-coil windings, the brushes should be so 
placed as to come into connection with the coils as they pass 
through the neutral regions between the poles. 

The coil PO is about to pass under a south pole and will 
soon reach the position of L K^ in the figure, after leaving 



Digitized by VjOOQ IC 



28 DYNAMOS AND DYNAMO DESIGN §13 

which it will pass under a north pole; but before doing this, 
it must again have its current of 50 amperes reversed. 
From the arrows in the diagrams it is seen that when the 
position L K\s reached, the segments must come in contact 
with another brush b and deliver thereto 100 amperes. In 
general it may be stated that in any closed-coil winding, in 
order that the current in the coils shall be reversed as the 
coils pass through every neutral region between poles, the 
segments in which they terminate must come in electrical 
contact with a brush and current must either be taken from 
or delivered to the winding by the brush. 

36. The neutral points are of two kinds NS or SNy and 
since these must alternate the signs of the brushes also alter- 
nate. In Fig. 26, it will be noticed that brushes a^ c^ and e 
are negative and brushes b, d, and /are positive. Brushes 
of the same sign are at the same potential and are shown 
connected together. They are at the same potential because 
the poles of the field magnet are alike in size and strength 
and there are as many coils from a to ^ as from b to r, each 
developing the same E. M. F., hence the total E. M. F. 
developed by the coils a, b is equal to that of the coils ^, c in 
value but of opposite sign, since they are under dissimilar 
poles, so a and c are at the same potential. In the same way 
c and e are at the same potential so that the negative brushes 
are all at the same potential. In the same way it may 
be seen that the positive brushes also are all at the same 
potential. 

36. In the winding shown in Fig. 26, the current enters 
by the brushes a, c^ and e, dividing at each brush, as shown 
by the arrows, and going by six paths, or circuits, from the 
negative to the positive brushes. If any one brush were 
omitted, two of the paths would not be supplied with cur- 
rent and the remaining ones would have to take more than 
their share. This would cause an undesirable unbalancing, 
and parallel or multicircuit windings are therefore always 
provided with as many brushes as there are neutral points; 
since there are as many neutral points as poles, there are 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 29 

as many brushes as poles. Further, it will be seen that such 
a winding as shown has as many paths as there are brushes, 
or, it may be stated, has as many paths as there are poles. 

37. This type of winding has many paths and is much 
used for small, low-potential machines with comparatively 
large current outputs, as well as for all large dynamos of 
great current output. The numerous paths offer low resist- 
ance to the currents, making the heat developed in the 
windings due to their resistance small, and also keeping 
down the size of the conductor required on the armature. 

Another point to be noticed is that any number of coils 
whatsoever may be made up into this kind of winding. 
That this is so may be seen by considering a coil added or 
omitted in Fig. 26. If the new number of coils are respaced 
equally, the winding will still have all the properties referred 
to above. 

38. Reentrancy. — In Fig. 26, were we to start at any 
point in the winding, say at A, and follow it around to B, 
to Cf and so on, we would eventually return to A ; or, in 
other words, the "winding closes or comes back on itself. It 
is from this property that the winding is called a closed-coil 
winding. Another way of expressing the same idea is to say 
that it is a reentrant uvlndlng. If we enter the winding 
at A and follow it around until we come again to A, the 
winding is said to reenter. If in following the winding 
around we traverse all the armature coils before reenter- 
ing, the winding is said to be singly reentrant. 

In following the winding. Fig. 26, around from A to B^ 
B to C, etc., it will be noticed that all coils are traversed in 
succession without skipping any, and the commutator is 
encircled but once. Because of this, the winding is called a 
slngrle winding. It should be evident that single wind- 
ings are necessarily singly reentrant. 

39. Summary. — The full name of a winding in which 
coils terminate in adjacent segments, as shown in Fig. 26, is 
single parallel winding, also often called a single multi- 
circuit winding. It is singly reentrant, may be formed of 



Digitized by VjOOQ IC 



30 DYNAMOS AND DYNAMO DESIGN §12 

any number of coils whatsoever, have as many paths or cir- 
cuits from — brushes to + brushes as there are poles, and 
must be provided with as many brushes as there are poles. 

40. Double Winding:. — Suppose that between every two 
commutator segments of Fig. 26, other segments are inserted 
and another similar winding connected to these segments. 



PIO. 27 

Fig. 27 shows such a winding ; it will be noticed that each 
coil ends in segments separated by an intervening segment. 
The current entering the winding from, say, negative 
brush Uy will divide between the segments O and 15, From O 
it goes through the coil O N-N M-L K, which is in contact 
with positive brush b. Again, from a to segment 15 it goes 
through coil 15 H-U lS-13 12-12 11, which again is in 
contact with positive brush b. 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 31 

41. It will thus be seen that there are now two paths 
between brushes a and b^ while in Fig. 26 there was but one; 
and also two paths between a and /, c and d, c and d^ e and d^ 
and e and /*, twelve in all, so this winding has twice as 
many paths as poles. The brushes used with this winding 
must be thick enough to always connect to two segments, 
for if brush a^ say, only connected to bar O^ the two paths 
from 15 would be open-circuited and would not be supplied 
with current. 

42. Investigating the reentrancy of the winding as 
before, starting, say, at segment A^ and following it around 
to B-Cy etc. to W-X, when A. is reached the winding closes. 
Only each alternate one of the segments and coils has been 
included and we have encircled the commutator once. In 
order to include all the coils it is necessary to again enter, say, 
at segment i, and follow around to 2-8-1^^ etc. to ^^ and back 
to i, when this second winding reenters. Such a winding is 
said to be doubly reentrant. The commutator has been 
encircled twice, so that the winding is therefore said to be a 
a double nvlndingf ; in fact, it is obviously a double wind- 
ing, since it consists of two interlaced single windings, each 
exactly like that shown in Fig. 26. 

43. It has been stated that a single parallel winding 
may be wound with any number of coils whatsoever; hence, 
each of the windings in Fig. 27 may consist of any number 
of coils. However, since both windings must have the same 
number of coils, it follows that doubly reentrant, double 
parallel windings similar to Fig. 27 may be formed from any 
even number of coils. 

44. Slngrly Reentrant, Double Winding. — In Fig. 28 
is shown a double parallel winding for four poles with an 
odd number of segments and coils. It is a double winding 
because all the coils terminate in segments removed from 
one another by one; hence, in following the winding around 
it will be necessary to encircle the commutator twice before 
all the coils are included. Starting at segment A and 
following around to B-C-D^ etc. to M-N^ it is found that the 



Digitized by VjOOQ IC 



32 



DYNAMOS AND DYNAMO DESIGN 



§12 



commutator has been encircled but the winding does not 
close, and, continuing, it goes to 1-2-3-4^ etc. to 12-13-A, 
closing only after encircling the commutator twice. It is 
therefore a singly reentrant, double, parallel nvindingr. 
It will be found to have eight paths from negative to posi- 
tive brushes or twice as many as there are poles. 




FlO. 98 

By omitting two coils in the winding shown in Fig. 28, 
the resulting winding will be of exactly the same character 
in every respect so it may be stated generally that to obtain 
this winding there must be an odd number of segments and 
coils. 

46. Summary. — Double parallel windings are those in 
which the coils terminate in segments adjacent but one. 
There must be as many brushes as poles and each must be 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 33 

thick enough to always touch two segments at once. The 
winding has twice as many paths as there are poles and any 
number of coils whatsoever may be connected up into this 
winding, but if the number of coils is even, the winding will 
be doubly reentrant, if odd, it will be singly reentrant. 

46. Triple Windings. — Triple ipvlndlngrs are very rarely 
used, as the action of the commutator is not good on account 
of the very thick brushes necessary. However, their prop- 
erties may be determined in the same manner as those for 
single and double windings and may be stated as follows: 

Triple parallel windings are those in which the coils ter- 
minate in segments adjacent but two. There must be as 
many brushes as poles and each must be thick enough to 
always touch at least three segments at once. There are 
three times as many paths as there are poles and any number 
of coils may be connected up into this winding. If the 
number of coils is divisible by three, the winding will be 
triply reentrant ; if not, it will be singly reentrant. 



8EBrES-WrNl>lN09 

47. Another type of armature windings is called series, 
or two-circuit, in distinction from the parallel, or multi- 
circuit, type already described. It is a peculiarity of the 
closed-coil windings as distinguished from the open-coil type 
that there are many coils in series on all paths between 
brushes. This amounts to the same thing as saying that 
each coil generates but a small fraction of the voltage of the 
machine, since this last is only the voltage of a single path, 
because the paths are connected in parallel, consequently 
their voltages cannot be additive. 

48. Referring to Fig. 26, suppose each coil, except those 
under commutation, develops 10 volts and also suppose that 
the negative brushes are at zero potential. Segment -P will 
be at zero potential, being in contact with a negative brush, 
Q will be at a potential of 10 volts, R at 20 volts, 5 at 
30 volts, which would be the potential of the machine as it 



Digitized by VjOOQ IC 



84 



DYNAMOS AND DYNAMO DESIGN 



13 



is in contact with positive brush /, so T also will be at 
30 volts, U at 20, etc. These potentials are marked on 
Fig. 29, which shows the commutator of the winding, Fig. 26. 
It is necessary that the E. M. F. between adjacent segments 
be kept low ; windings in which the potential does not vary 
approximately uniformly from segment to segment, as shown 
in Fig. 29, are therefore undesirable. It will be noticed 




Pig. 39 



that the gradations from ^ to ^ are similar to those from 
^ to ^ or from e to /; so that a point between a and b will 
always be at about the same potential as a similarly located 
point between c and d. Points on the commutator that are 
about two poles apart will always remain at about the same 
potential, and, in fact, with parallel windings for large gen- 
erators it is quite customary to design the windings with a 



Digitized by VjOOQ IC. 



§12 DYNAMOS AND DYNAMO DESIGN 35 

certain whole number of segments for every two poles and 
permanently connect all segments exactly two poles apart 
to large equalizer, or cross-connecting:, rlng^s, which will 
be illustrated later. 

49. In the series-type of winding, the coils are connected 
to segments removed from one another by approximately 
the angle of two poles, and, assuming that the potentials on 
the commutator are to vary about as in Fig. 29, it will be 
seen that the ends of any coil will not be at any very great 
potential difference; that is to say, each coil will have to 
develop but a part of the E. M. F. of the machine. In fol- 
lowing such a winding around, an advance of two poles is 
made for each coil traversed, so that the commutator would 

be encircled once for every ^ coils, where/ is the number of 

poles. In series- windings, the number of times the commu- 
tator is encircled in following the winding out has nothing 
whatever to do with the number of windings, as it does in 
the case of the parallel type. In series- windings, the dis- 
tinguishing feature is the connection of the terminals of this 

series of ^ coils. 

60. Slngrle Series- Winding:. — A singrle series-Tvlndlnfir 

is one in which a series of ^ coils encircles the commutator, 

and the terminals of this series are connected to adjacent 
commutator segments. It is immaterial whether this series 
of coils makes a revolution plus one segment or less one seg- 
ment on the commutator. 

61. In Fig. 30 is shown a diagram of a single seri as- 
winding with six poles, so that a series of ^, or three, coils 

terminates in adjacent segments. It is not intended in the 
diagram to represent the coils themselves any more than in 
Figs. 26, 27, and 28, and the lines here shown inside the 



Digitized by VjOOQ IC 



36 DYNAMOS AND DYNAMO DESIGN §13 

commutator are simply to show where the ends of each arma- 
ture coil are connected. Complete diagrams of these wind- 
ings will be given later. 

62« The winding is much more complicated than a single 
parallel winding and the diagram must be studied very care- 
fully. Investigating the number of paths in the same 



S 



PlO. 80 



manner as before, we find negative brush n touches seg- 
ments £9 and SO. The coil extending from 29 to the right 
will be found to be connected to segment 8, from which 
another coil connects to J9 and to 30. This, then, is not a 
path, for both segments !20 and SO are in contact with brush a^ 
which therefore short-circuits the series of coils S9S, 8-19^ 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 8t 

19S0 at the instant shown. From 29^ however, another 
coil connects extending toward the left to 18-7-28-17-6- 
27-16-6-26-15-4-25-H to where S touches the + brush b. 
This, then, is one of the paths through the winding. 

Starting at a again and going to segment 80, another path 
will be found to connect to 9-20-81-10-21-82-11-22-1-12-23- 
2-18-24-8 and out. It will be noticed that every coil in the 
winding has been traversed and that there are but two paths. 
In fact, the single series-winding has but two paths regard- 
less of the number of poles. 

63. It will be noticed that in Fig. 30 but two brushes 
have been shown in full lines — one positive and one nega- 
tive — while at the other neutral points they are shown dotted. 
In following the winding from segment 29 to 8, 8 to iP, 19 to 
SO, it will be seen that this series of coils has segments 29 
and 80 under negative brush a, segments 8 and 9 under 
negative brush c, and segment 19 under negative brush e. 
Now it has been stated that the segments to which the coils 
connect should either be supplied with current or have cur- 
rent taken from them by a brush in order that the current 
in the coils shall be reversed as the coils pass through the 
neutral regions, and it is easily seen that the single negative 
brush a is sufficient to accomplish this for the series of coils 
under consideration without the use of the brushes c or e. 
However, the coWs 29-8, 8-19, 19-80 are, at the instant shown, 
not developing any E. M. F. , being in the neutral region, so the 
segments to which they connect are practically at the same 
potential: other brushes c and e may be put on the commu- 
tator and permanently connected to a if the brush a is not 
sufficiently large to carry satisfactorily all the current of 
the armature. 

64. winding Requirements. — If we start at any point 
in the winding, say at 1, and follow it around, we find 
that the coil connects segments 1 and 12, or it might 
be said that the ends of the coil span 11 segments 
on the commutator. The next coil to this is 12-28, which 
is 11 coils in advance of coil 1-12, It would be said of this 



Digitized by VjOOQ IC 



38 DYNAMOS AND DYNAMO DESIGN §12 

winding on this account that the pitch was 11. It will be 
observed that the number of segments spanned by a coil 
must always be equal to the pitch of the winding numeri- 
cally. Now since a series of ^ coils encircles the commu- 

tator with one segment more or less, and since each coil 
spans a certain number of segments, called the pitch, it fol- 
lows that the only numbers of segments that can be used for 

such a winding will be ~ 1^ ± 1, where Y is the pitch, 

being any integer. If C is the number of coils, then C 
also is the number of segments, so that the number of 

coils = C=:Iy ± 1. In Fig. 30, C = 32 = J X 11 - 1. 

Where the minus sign is used it means that a series of ^ coils 
spans one more segment than a revolution, while the plus 
sign indicates that ^ coils lack one segment of spanning a 

complete revolution. 

If the winding is followed through from, say, segment 1 
to i^, 23, 2, etc., it will be found that before returning to the 
starting point all the coils and segments will have been 
traversed ; it is therefore a singly reentrant winding. 

56, Summapy. — A single series-winding is one in which 
the ends of each coil embrace an angle on the commutator 
about equal to twice the angle between the centers of poles 

and in which a series of — coils is connected to adjacent com- 

mutator segments. It has but two paths or circuits, regard- 
less of the number of poles, and two brushes, one negative 
and one positive, are all that is required, although all 
the neutral points on the commutator may be used for 
brushes if desired. It is singly reentrant, and the number 
of coils possible for this winding must conform to the follow- 
ing formula 

C=|f±1. (1) 



Digitized by VjOOQIC 



§12 DYNAMOS AND DYNAMO DESIGN 3fl 

where C = number of coils; 
/ = number of poles; 
Y = any integer, being the pitch of the winding. 

66. Double Series- Windings. — In the same way as 
with parallel windings, we may double the number of seg- 
ments and coils of Fig. 30 and connect them up into two 
separate' windings, each exactly like Fig. 30. Such a wind- 
ing would be a double winding and would be doubly reen- 
trant, since it consists of two separate single windings. 
Bach winding would have two paths, so that there would be 
four paths for the two windings. However, in order that 
all the paths may be available, it will be necessary to have 
brushes thick enough so that at least one of each polarity shall 
always be in contact with some point in each winding. In 
regard to the^ number of coils, it will be noticed that a series 
of coils end on segments adjacent but one, because between 
every two adjacent segments we have put in a new segment 
belonging to the second winding. The new winding would 
thus have 64 coils and the number of segments would have 
to conform to the formula 

r=|K±2 (2) 

The value of Y is here 22, as it is double the value in the 
winding. Fig. 30, and hence, 

C=|x22-2=64 

67. Sammary. — In general, double series-windings are 

those in which a series of ~ coils terminates in segments 

adjacent but one. The winding will be doubly reentrant if 
half of the segments and coils can be connected into a single 
series- winding. Dividing formula 3 by 2 gives 

2 ^ 2^2^^ 

Comparing this with formula 1 for C, for the single 

C 
series-winding, it will be seen that this value of — will form 



Digitized by VjOOQIC 



40 DYNAMOS AND DYNAMO DESIGN §12 

Y 
a single series-winding if ~ is an integer, or, in other words, 

if Y is an even number; if Y is odd, the winding will be 
singly reentrant. 

From what has been shown, it can be said that a triple 

series- winding: is one in which the series of ^ coils termi- 

nate in segments adjacent but two. The number of coils 
possible must satisfy the formula 

C = {y±Z (3) 

The winding will have six paths regardless of the number 
of poles, and while two brushes are all that is necessary, they 
must be thick enough to touch at least one segment of each 
winding at all times. The winding will be triply reentrant 
if Y is divisible by 3 without a remainder, if not, it is singly 
reentrant. 

68, Single windings are very generally used, double 
windings, both parallel and series, are occasionally found, 
while triple and quadruple are seldom if ever used in prac- 
tice today. It often happens that a number of different 
windings are possible with a certain number of segments and 
bars (which is a great advantage, as will be explained later, 
for the same armature winding, so far as coils and commu- 
tators are concerned, may be used for several voltages). 
By using different commutator connections, the different 
windings may be made to give different voltages. For 
instance, the winding, Fig. 29, for six poles thirty-two coils 
and bars may be connected info a single or double parallel 
winding, since any number of coils may be used for these 
windings. It can also be connected as a single series-wind- 
ing, and since 32=C=|XlO + 2it can also be used for 
a double series-winding. All six-pole windings have this 
advantage. 

Singly and doubly reentrant windings do not differ at all 
in operation and there is no special advantage with either; 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 



41 













4-» cn 






CO 


CO a; 




CO 


>* 
















3 "»-» 






>^ 










to 














^ c 








^.1 




^ 


Xi 


-H 








Ji 






^ E 






VM > 


a; 


M-l 


V 


^ 








o* 






C bo 








a; 














6- 






0) o 
o en 

rt o 






sign must 
winding 




^4 

x; 

4-> 




3 
*> 


^ II 






••- 


_... 






|2 


CO 




C/3 


>> 

X 




'5 


•5 ^ 










4^ 






CO 


.J2 












a 








_Q G 














M 






c 












_c -^ 




^ 






-H 


a; 




1 •> 


4) 

3 






acent 
esegm 


H 




of eac] 
t of eac 




VM 

3 


G 

0) 




II 


to 

3 

1 

4-* 

3 


















ne brush 
segmen 


C/5 


1 


> 
0) 




^J 














3 


















least o 
ouch a 


"3) 








II 


1 


















*j ^ 


G 








'o 

o 




















< 


. c2 


^ 










H 
5 


bo 


o ^ 

G ^ 




CO 

*> 


« 
£ 






H 














1. 


in 


3 




*'5 


4-» 






^ 














« CO 




VM 


>> 






h) 














^ 


a; S 


>>-Jo 

bO"> 


V 


^ 


-Q 






P; 
















i s 




[a. 


^ 


>> 




^ 


















^ ;;; 


x: 


)^ 


3 


G 




H 


* 
c 

c 




< 


<4-> 




e 


w! 


•c2 ^ 


4-> 


e 




< 






a; 

o 


4-> 

G 
1) 


G 

i 

bo 




G 

S 


CO 

a; 


CO 
4-> 

G 

0) 

E 

bo 


' if odd 
y if even 










S8 

a 

03 




5 





S 

•^^ 






0) 


a 


CO 

o 


•^1 








>> 






a! 




T3 


s 




'? 


CO 


^ 


c 5 








G 


? 






<J 


o 




^ 


ctf 


H 


CO Q 








<J 


5 


















4-> 












>* 






V 


4-> 


CA 

G 








G 

E 












' c 






C 

a: 


c 
o 

<5 


6 

bo 




G 

E 
to 


to 
a; 

1 


bo 

c 
O 


"bb 
G 

c2 








G 






































c 

4-> 




1 c 


4-> 




• 4-> 

-^ -G 


8 
.E 








8'" 


* 








C 




-'^i^ 'Z 


o. 




|| 


.. . 4-> 








0) 

VM -^ 










6 




"S rt 


^-1 

o 




e 2 










® -^ bo 










4-> 










s « 

s: >> 










umber 
poss 
windir 










CO 

1 




s s 


B 

3 




CO rt 

2 ^ 


«« e ^. 


















<j 


» 




m 


^ 








a 1 





Digitized by VjOOQ IC 



4a DYNAMOS AND DYNAMO DESIGN §12 

the reentrancy of a winding is simply a matter of interest 
to the student. 

69. It will be noticed with regard to the restrictions as 
to the numbers of coils in the single series-windings that 

if ~ is even C must be odd or else the winding is impossible, 

but if ^ is odd then C may be either odd or even according 

as Y is even or odd. For four poles or eight poles, any odd 
number will do, while for six poles any number of coils not 
divisible by three is possible. For eight, ten, or more poles 
series-windings are rarely used and thumb rules are a bur- 
den rather than a help. 

60. The important peculiarities of the various windings 
that have been considered are collected for comparison and 
reference in Table I. 



THE MAGNETIC CIRCUIT 

61. As far as the generation of the E. M. F. of the 
dynamo is concerned, it is only essential that the. lines of 
force of the magnetic field be present at the points where 
they are cut by the conductors, and have the proper direc- 
tion and distribution. However, since each line of force is 
continuous, forming a closed circuit, provision must be 
made for a complete path for the lines of force to and from 
the points where they are cut by the conductors, and through 
the magnetizing coil or coils wherein they are generated. 
Of course, they might be left to find their own circuit 
through the surrounding air, but in order to obtain the 
large number of lines of force required with the expenditure 
of a reasonable amount of magnetizing force, it is necessary 
that the path of the lines of force be of as great a perme- 
ability as possible; i. e., through an iron or steel magnetic 
circuit. 

In addition to the armature and its winding, a bipolar or 
multipolar dynamo must have an iron or steel frame, or 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 43 

field magnet, which completes the magnetic circuit outside 
the armature. This frame is made up of one or more pairs 
of pole pieces, from (or into) which the lines of force pass to 
(or from) the armature through the spaces between the 
faces of the pole pieces and the surface of the armature 
core, which are called the air gaps ; it must also have a part 
upon which the magnetizing coils are wound, which part is 
called the field core. The part of the frame that joins 
together the field cores, if more than one is used, or that 
joins the pole pieces and the field cores, is called the magr- 
netic yoke. 

62. It will be seen that the object of the frame, as a 
whole, is to so guide the lines of force that are set up by 
the current in the magnetizing coils that they will enter and 
leave the armature at the proper points, forming the mag- 
netic field in the air gaps of the required distribution and 
density. 

It is not essential to the operation of the machine that the 
frame be of any given form or size, so long as the lines of 
force are properly delivered to the armature; economy in 
materials or labor, mechanical strength, and other consider- 
ations determine the form and size of frame to be adopted. 

63. Since the magnetic circuit may be considered anal- 
ogous to an electric circuit, it will be seen that in order to 
obtain a large number of lines of force with a moderate 
magnetizing force, the reluctance of the circuit must be 
low; that is, the iron should be of considerable cross-section 
and the circuit of moderate length. It should be remem- 
bered that, since the permeability of the best of iron is only, 
perhaps, 1,500 times that of air, a considerable number of 
lines of force that pass through the magnetizing coil com- 
plete their circuit around through the air without passing 
through the air gaps. To reduce this magnetic leakage as 
far as possible, surfaces between which there is a great dif- 
ference of magnetic potential should be kept as far apart as 
the design of the magnet will allow, and made of as small 
area as possible. In any case, some leakage is bound to 



Digitized by VjOOQ IC 



44 DYNAMOS AND DYNAMO DESIGN §12 

occur, and this must be provided for by making^those parts 
of the frame through which the leakage lines pass of suffi- 
cient area for both the useful and the leakage lines. The 
conditions that govern the leakage will be more fully dis- 
cussed later; in general, the area of the iron in the frame 
must be sufficient for from 10 to 50 per cent, more lines of 
force than are used in the armature. 



DEN8ITT OF liLNlES OF FORCE 

64. Referring to the saturation curves for iron and steel, 
The Magnetic Circuity it will be seen that the saturation 
curves there shown rise in a nearly straight line for some 
distance from the origin, then curve away from the axis of 
the ordinates and follow another approximately straight 
line, which makes a much greater angle with the axis of the 
ordinates than does the first-mentioned line. This effect is 
much more marked in the case of wrought iron and cast 
steel than with cast iron, but in any case it will be seen 
from this feature of the saturation curves that the most 
economical density at which to work the iron of the mag- 
netic circuit is that in the vicinity of the bend, or ** knee," of 
the curve. A much lower density could not be economically 
used, because a considerable increase in the number of lines 
of force could be obtained with comparatively little increase 
in the magnetizing force required; and on this account 
accidental small changes in the magnetizing force would 
produce a considerable change in the number of lines 
of force, so that the magnetic circuit of the machine would 
be in an unstable condition. A much higher density would 
not be economical, because the increase in the number of 
lines of force could be obtained only by a very considerable 
increase in the magnetizing force. 

65. Applying these statements to the curves just men- 
tioned, it will be seen that, in general, cast-steel and wrought- 
iron forgings should be worked at densities of between 80,000 
and 100,000 lines of force per square inch, while sheet iron 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 46 

may be worked higher — between 90,000 and 110,000 lines of 
force per square inch. With cast iron, the curves being 
flatter, the allowable range is somewhat greater, the usual 
range in practice being from 25,000 to 50,000 lines of force 
per square inch, the latter value being used only in the case 
of the best grades of soft, gray cast iron. 

The best densities to use are, therefore, not those that 
give the maximum permeability of the iron used, as at that 
point the iron would be in the unstable condition referred 
to previously. 

66. From the foregoing, and from the curves referred to, 
it appears that for the same expenditure of magnetizing force 
a cast-iron magnetic circuit must have about twice the sec- 
tional area of one of cast steel or wrought iron, in order to 
obtain the same number of lines of force, so that the cast- 
iron magpetic circuit will be about twice as heavy as one of 
steel or wrought iron. As it costs considerably less per 
pound, however, this extra weight is often counterbalanced; 
in fact, the choice of materials for the frame, as well as 
almost all the other features of a dynamo, depends on the 
special conditions governing each particular case. 

67. The density used in the air gaps varies, but the best 
practice fixes it somewhere between 30,000 and 70,000 lines 
of force per square inch, depending on the design. In any 
case, the amount of the magnetizing force that is required 
to force the magnetic flux through the air gaps is a large 
proportion of the total amount, since the permeability of 
the air gaps is 1, which much more than compensates for 
their comparatively short lengths. 



FORM OF MAGNETIC CIRCUIT 

68. The form of the magnetic circuit is subject to many 
variations; there are, however, two general classes into 
which they may all be divided. In the first, a single source 
of magnetizing force for each pair of poles (which may 
reside in one or more magnetizing coils) sends the lines of 



Digitized by VjOOQ IC 



46 



DYNAMOS AND DYNAMO DESIGN 



§12 



force around through a magnetic circuit, of which the air 
gaps and armature directly form a part. Such an arrange- 
ment is said to have salient poles. In the second type, 
at least two magnetizing forces are necessary for each pair 
of poles; these magnetizing forces act in opposite directions 
upon a complete magnetic circuit, and the opposing lines of 
force cause poles to appear at points on the magnetic circuit, 
which points are properly provided with pole pieces, between 
which the armature is located. Such an arrangement is said 
to have consequent poles. 




PlO. 81 



69. One of the simplest forms of salient-pole bipolar 
field magnets is represented in Fig. 31. In this form the 

magnetizing force is supplied by the 
single coil shown in section at IV^ W. 
This surrounds the field core C, 
to which are attached the magnet 
yokes J/, M, which terminate in the 
pole pieces A\ 5, between which 
the armature A revolves. The mean 
paths of the lines of force through 
the magnetic circuit (neglecting 
leakage lines) are indicated by the 
dotted lines having the arrowheads, 
which indicate the din^^.ction of the lines of force, assuming 
the polarities of the pole pieces to be as indicated by the 
letters TV, S. In this figure the field core is represented as 
being vertical; this type of magnet is so used in certain 
machines of English make. It may, however, be either ver- 
tical or horizontal, and be above, below, or on either side of 
the armature, as desired. The Jenney motors, the Wood 
bipolar machines, the Holtzer-Cabot small motors, and 
others made in this country have used this type of magnets 
with the coil horizontal and below the armature. Further, 
the armature shaft may either have the direction indicated 
or be at right angles to that direction, if desired, without 
changing the character of the field magnet. The mechanical 
construction in this last case would evidently be bad, and this 



Digitized by VjOOQ IC 



13 



DYNAMOS AND DYNAMO DESIGN 



47 




PIO. 8S 



is generally the principal feature that determines the dispo- 
sition of the magnet frame with regard to the armature. 

70. A form of consequent-pole field magnet derived from 
that just described is shown in Fig. 32. This form is known 
as the Manchester type^ 
and has been used by 
the Mather Electric 
Company, the Westing- 
house Company, and 
others in this country. 
It is practically the 
same form of magnet 
as that shown in Fig. 31, 
with the addition of a 
second similar magnet 
situated on the opposite side of the armature A^ as indicated 
by the letters N\ M\ C, M\ and S\ 

Assuming that the same total number of lines of force 
passes through the armature in each case, it follows that 
with the consequent-pole magnet. Fig. 32, each half of the 
magnetic circuit contains half the total number of lines, and 
needs, therefore, to be of but half the sectional area of the 
frame of the salient-pole magnet, which carries all the lines 
of force, as is indicated by the relative proportions of the 
two magnets. (See Figs. 31 and 32.) Consequently, the 
weight of the frame in either case is about the same. 

71. In the consequent-pole magnet, the magnetic circuit 
in each half is approximately the same length but of half 
the area as that of the salient-pole magnet ; its reluctance is 
about twice as great, but since it carries half the number of 
lines of force, it follows that the magnetizing force required 
for each half of the consequent-pole magnetic circuit is the 
same as that required for the whole of the salient-pole mag- 
net. However, the magnetizing coils on the consequent- 
pole magnet are of smaller diameter than those used in the 
salient-pole magnet, so that the weight of copper used for 
the magnetizing coils of the former type of magnet is not 



Digitized by VjOOQ IC 



DYNAMOS AND DYNAMO DESIGN 



§12 



double that required for the latter type. The actual ratios 
of weights of copper and iron may be readily calculated for 
any particular case, but there are other conditions that influ- 
ence the choice of the form of magnet to be used, which 
must be taken into account. 



72. Fig. 33 shows the adaptation of these two forms of 
field magnets to a multipolar machine. In the figure, the 

part to the left of 
the vertical diam- 
eter represents the 
salient-pole magnet, 
and that to the right 
represents the conse- 
quent-pole magnet, 
each being laid out 
as for an eight-pole 
magnet. 

The salient-pole 
magnet consists oi a 
number of separate 
magnets, each with 
its magnetizing coil. 
It is, therefore, nec- 
essary to supply 
some separate sup- 
port for these magnets. In the consequent-pole magnet, 
however, the whole frame is continuous, each pole piece 
being supported by a field core on each side, the frame, 
therefore, being of sufficient mechanical strength for its own 
support. In the latter form, the mean length of the 
magnetic circuit for each pair of poles is less than with the 
salient-pole magnets, which results in a slight saving in 
magnetizing force, other things being equal. 

Of the above types of magnets for multipolar machines, 
the salient-pole type has been used in the Perrett machines, 
built by the Elektron Manufacturing Company, and the 
consequent-pole type has been used by the Standard Electric 




FIG. 88 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 



49 




PIO. 84 



Company, in this country, and in several types of machines 
made abroad. 

73. The two simple forms of field magnets that have 
been described may be considerably modified by changing 
the position or increasing 
the number of the field coils. 
For example, the magneti- 
zing coil of the salient-pole 
magnet. Fig. 34, may be 
wound over the entire frame 
from pole piece to pole piece, 
as in the old style ring-type 
machine of the Mather Elec- 
tric Company. Similarly, 
the magnetizing coil on each 
half of the consequent-pole 
magnet. Fig. 35, may be wound over the entire frame from 
pole piece to pole piece, as in the C & C machines. In both 

these examples, the field cores are 
made approximately circular in out- 
line. 

Further, by dividing the magnet- 
izing force between two coils, and 
locating these coils in the part indi- 
cated as the magnet yoke J/, My 
Fig. 31, a type of field magnet re- 
sults that is commonly known as 
the horseslioe type, as illustrated 
in Fig. 34. It will be seen that in 
these two forms the magnet yoke M 
of each corresponds to the field core 
of the other. This type of field mag- 
net is very extensively used for 
bipolar machines, the Thomson- 
Houston, Crocker- Wheeler, Key- 
stone, and other makes of machines 
using it in the position shown, i. e., with the magnet frame 
beneath the armature. 

44—5 




PlO. 85 



Digitized by VjOOQ IC 



50 



DYNAMOS AND DYNAMO DESIGN 



§12 



The General Electric Company, in their Edison machines, 
the Eddy Electric Manufacturing Company, and others have 
used the same form of magnet in the reverse position, i. e., 
with the magnet frame above the armature. 

The Excelsior arc machine employs the same type of mag- 
net, but with the armature shaft parallel to the field cores, pass- 
ing, therefore, directly through the magnet yoke. The pole 
pieces are necessarily modified in shape to suit the changed 
position of the armature, and are extended to embrace the 
three outside faces of the armature, which is ring wound. 

74. The consequent-pole magnet that results from com- 
bining two horseshoe magnets of the types illustrated in 
Fig. 34 is shown in Fig. 35. Here the various letters have 
the same reference as in the previous figures. As in that 
previously described, the consequent-pole arrangement 
requires only half the cross-section of metal in each half of 
the magnetic circuit, but the total amount used is about the 
same. This is also a commonly used type of bipolar field- 
magnet. Among others, it is used in the Wood arc machine 
of the larger sizes, in the position represented in the figure, 
i. e., with the field cores C, C, C, C vertical. The Weston 
machine used the same form of field magnet, but with the 
field cores horizontal; it has also been used in this same 
position for various special machines built by the General 
Electric Company and others. 

The smaller sizes of the Wood arc machine use this form 

of magnet with the field 
cores horizontal, and with 
the shape of the pole pieces 
modified so as to allow 
the armature shaft to be 
parallel to the field cores, 
it passing through and 
having its bearings in the 
yokes M, M. The Brush 
arc machine uses a simi- 
lar construction, but the 
armature is made in the form of a ring- wound disk with the pole 




MMMMT. 



PlO. 36 



Digitized by VjOOQ IC 



§13 



DYNAMOS AND DYNAMO DESIGN 



51 



faces toward the end faces of the armature, as represented 
in Fig. 36. The magnet in this case might be considered to 
be two separate bipolar, salient-pole, horseshoe magnets. 

75. By carrying the magnetizing coils still farther along 
the frame, until they are as close as possible to the ends of 
the pole pieces, still another type of field magnet results, as 
represented in Fig. 37. As shown, this is a very heavy and 




FIO. 87 

clunisy magnet, requiring a large amount of material on 
account of the length of the magnet yoke J/, M. If, how- 
ever, half the material in this yoke is located on the other 
side of the armature, so that the magnetic circuit through 

the frame from field core ^ 

to field core consists of i ^^K^. 

two branches, a much 
neater and lighter mag- 
netic circuit, which is 
quite extensively used, 
results, as represented 
in Fig. 38. 

This form of circuit 
still has salient poles, 
since the poles are pro- 
duced by the direct 
action of the magnetizing forces and not by the opposition 
of two magnetizing forces. It has the advantage that the 




Fig. 88 



Digitized by VjOOQ IC 



62 



DYNAMOS AND DYNAMO DESIGN 



la 



magnetizing coils and armature are enclosed by the frame, 
thus aff ord'ing them mechanical protection. 

The Thomson-Houston arc-lighting dynamos employ this 
type of field magnet, the form being modified by making the 
magnet yokes of a series of round, wrought-iron bars, which 
connect together circular flanges on the ends of the field 
cores, thus making the general outline cylindrical. Eicke- 
meyer used it for very compact machines in which the mag- 
netizing coils actually enclose the armature, the field cores 
being very short. 

The same form of magnet, but with the magnetizing coils 
above and below the armature, was used in the old Hoch- 
hausen dynamos, also by the Thomson-Houston Company 
for their old ** S. R. G." railway motors, and by others. 

76. With this arrangement of the magnetizing coils, a 
consequent-pole bipolar magnet is not possible; but by 
^y^ ]f X reversing one of the coils 

^ /' .-_-__-^N so that the two mag- 
netomotive forces are 
opposite, two consequent 
poles will be formed on 
the magnet yokes M^ J/, 
Fig. 39, at a point oppo- 
site the neutral spaces of 
the bipolar form ; and by 
locating suitable pole 
pieces at these points, a four-pole magnet results, as repre- 
sented in Fig. 39. It will be seen that this magnet has one 
pair of salient poles N^ N and one pair of consequent 
poles 5, 5. This gives a very compact form of four-pole 
magnet, and is used in several types of railway motors, 
in the Eddy slow-speed stationary motors, and by other 
makers. The Wenstrom dynamos also had a modified form 
of this type of field magnet, the magnet yoke being barrel- 
shaped and completely enclosing the magnetizing coils and 
pole pieces, spaces being left in the sides for the removal of 
the armature. 




Pig. 89 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 53 

77. By winding magnetizing coils around the conse- 
quent poles of the type of magnet illustrated in Fig. 39, 
they become salient poles, giving still another type of field 
magnet, illustrated in Fig. 40. The same letters of refer- 
ence are used in this figure as in the previous ones. This is 
a very useful form of field mag- 
net, and is the one most generally 

used in this country for multi- 
polar machines of any number 
of poles, almost every maker 
using it for multipolar genera- 
tors and alternators. 

The various magnet yokes 
form a complete ring, which is 
usually, especially when six or 
more poles are used, made circu- ^'o- ^ 

lar in outline. A modification of this form of magnet has 
been used by the Siemens & Halske Company, in which the 
field cores project radially outwards from a common hub, 
instead of inwardly, the armature revolving outside the 
poles of the magnet. 

78. The number of possible forms of field magnets is 
very great, although they may all be classed as either salient 
or consequent pole magnets, or combinations of the two. 
Many of the forms of magnets that have been and are used 
seem to have been designed merely with a view to getting 
something different from any other maker, and considera- 
tions of economy of material or of mechanical fitness, which 
should prevail in the selection of a design, have been largely 
neglected. These forms described are the basis of the 
designs of field magnets in modern construction. Nearly 
all modern machines are multipolar except in the small sizes 
or in the case of machines designed for exceptionally high 
speed. The type of field magnet that is by far the most 
commonly used is that shown in Fig. 40. This design gives 
an economical distribution of material, it can be easily 
adapted to any number of poles, and it also gives a machine 



Digitized by VjOOQ IC 



64 



DYNAMOS AND DYNAMO DESIGN §12 



of graceful outline. In some cases the pole pieces are cast 
with the yoke while in others they are bolted to the yoke. 
In most modern machines the pole pieces are laminated and 
are either bolted to the yoke or cast into it. 



METHODS OF EXCiriNG THE FIBIiB 

79. The requisite number of ampere-turns for exciting 
the field of a dynamo-electric machine may be obtained in a 
variety of ways. In the first place, the current that flows 
through the magnetizing coils may come either from some 
separate external source, the machine being then said to be 
separately excited, or it may be furnished by the arma- 
ture of the machine itself, it being then said to be self- 
exelted. In some cases a combination of separate and self- 
excitation may be used. 

A diagram illustrating separate excitation is given in 
Fig. 41. The current required is in this case supplied by 




Fio. 41 



the primary or secondary battery B^ although another 
dynamo may be used, if desired. In order to adjust the 



Digitized by VjOOQIC 



§12 DYNAMOS AND DYNAMO DESIGN 66 

current in the magnetizing coils to the proper value, or to 
vary it if necessary, an adjustable resistance r is included in 
the field circuit. 

The armature has no connection whatever with the field 
circuit, but supplies the external circuit Re directly. 

80. It is evident that with self-excitation a small or a 
large current may be used in the magnetizing coils, accord- 
ing to the nature of the source of the current, a large or a 
small number of turns being used in the magnetizing coils 
to give the necessary magnetizing force. 

Alternators are usually separately excited, since the cur- 
rent given out by the machine, being alternating, cannot be 
used directly for the purpose. Separate excitation has also 
the advantage that variations of the output of the armature 
of the machine, caused by changes in the speed or of the 
current, do not directly affect the field excitation. 

81. Cliaracterlstlc Curves. — In order to study the 
behavior of dynamos, it is instructive to lay out curves 
showing the relation between the current delivered by the 
armature and the E. M. F. at the terminals of the machine, 
or the E. M. F. generated in the armature. For example, 
in the case of the dynamo shown in Fig. 41, suppose the 
field is kept at a constant strength and the armature run at 
a constant speed. The external resistance Re is then varied 
so that the current supplied by the armature is varied 
in amount. If an ammeter is connected in circuit and a 
voltmeter connected across the brushes, we can, by vary- 
ing Re, take a series of readings and obtain the voltage read- 
ings corresponding to the various current readings. These 
points can then be laid off on cross-section paper, as shown 
in Fig. 42, and a line or curve a b obtained ; this curve is 
called the characteristic curve, or characteristic, of the 
dynamo. Volts are laid off as ordinates and amperes as 
abscissas. 

With a separately excited machine, it will be found that 
the voltage falls off as the current increases, as indicated by 
the drooping of ^ ^. • This falling off is due to two causes. 



Digitized by VjOOQ IC 



56 



DYNAMOS AND DYNAMO DESIGN 



§12 



In the first place, a portion of the total E. M. F. generated 
in the armature is used in forcing the current through the 
armature itself; hence, the voltage at the armature ter- 
minals is decreased by an amount equal to the voltage drop 
in the armature. The volts lost in the armature and at the 

























































• 




1_~" "— — 




--^ 


"»""■ 


..^, 














»^ 




■--. 




■~~i 












-w»- 


— i 
















0( 


9Exm 

00000 


m 


















D 


















1 








"(5)" 


) 












3 


'oLom 


r 
























- 








h 


^- 




' 




o 











larpju 


4 
IBB I 


1 




m 



Pig. 48 

brush contacts can always be found by multiplying the cur- 
rent by the resistance between the brushes, or if / is any 
given current and R^ the resistance between brushes, then 
the volts lost corresponding to the current / will be I Ra' 
At zero load it is evident that the voltage at the brushes 
will be equal to the voltage O a generated in the armature, 



Digitized by VjOOQ IC 



§1^ DYNAMOS AND DYNAMO DESIGN 57 

because there is then no drop in the armature. Another 
cause of the falling off in voltage is the weakening of the 
magnetic field by the current flowing in the armature. This 
will be fully explained later on in connection with armature 
reaction. For the present it will be sufficient to state that 
when current is taken from an armature, the magnetizing 
action of the armature currents is to a certain extent opposed 
to the original field and weakens it somewhat. This reduces 
the total E. M. F. developed in the armature, and this in 
turn causes a falling off in the E. M. F. at. the brushes. If 
it were not for the demagnetizing action of the armature, 
the field strength in a separately excited dynamo would 
remain constant, because the current flowing around the 
field winding is constant. 

82. The line ^^ is often called the external charac- 
teristic of the separately excited dynamo, because it shows 
the relation between the load or current and the external 
voltage. If the relation between the current and the inter- 
nal voltage, i. e., the voltage actually generated in the arma- 
ture, is to be represented, the armature drop or loss in voltage 
must be added to the ordinates of the curve a b. This is 
easily done as follows: Select any value of the current, say 
the number of amperes corresponding to the distance Od^ and 
multiply this current by the resistance of the armature 
between brushes. Lay off this value of the drop or lost volt- 
age, as shown by de. For example, if Od w&vg^ 60 amperes 
and the resistance between brushes .05 ohm, then de would 
correspond to 50 X .05 = 2.5 volts. Now the armature drop 
increases directly as the current, so if we draw a line O ef 
through the points O and e, the vertical distance between Ox 
and this line will represent the drop corresponding to the 
current. For example, the armature drop corresponding to 
the current O g \s g h. In order, then, to obtain the curve 
representing the total E. M. F. generated in the armature, 
the ordinates of the line Oe f are added to those of the 
curve ab^ thus giving the dotted curve akl. The dotted 
curve is sometimes called the total cliaracteristic of the 



Digitized by VjOOQ IC 



68 



DYNAMOS AND DYNAMO DESIGN 



13 



machine because it shows the relation between the current 
and the total E. M. F. developed in the armature. 

83. In a separately excited dynamo, therefore, the volt- 
age falls- off as the current is increased. The amount of 
the falling off will depend on the design of the machine, and 
the lower the resistance of the armature, the better will be the 
voltage regulation. If the armature resistance is very low, 
a separately excited dynamo will fall off but little in voltage 
as the load is applied, and if it is necessary to keep the volt- 
age absolutely constant, it can be done by increasing the 
field current. 

SEIirES- WINDING 

84. There are three general methods by which self- 
excitation is accomplished. In the first, the whole of the 

current flowing through 
the armature also flows 
through the magnetizing 
coils; such a machine is 
said to be serles->voiind, 
from the fact that the 
armature and magneti- 
zing coils are connected 
in series. This arrange- 
ment is represented in 
the diagram shown in 
Fig. 43. 

With this arrangement, 
the magnetizing force 
acting on the magnetic 
circuit, consequently the number of lines of force in the 
magnet, varies with the current that the machine furnishes 
the external circuit; therefore, when the armature is running 
at a constant speed, the E. M. F. that is generated in it varies 
as the current varies, though not necessarily in the same pro- 
portion. This is not usually desirable, since most applica- 
tions of direct current require that either the E. M. F. or 
the current be maintained approximately constant. 




Pig. 48 



Digitized by VjOOQ IC 



§1^ 



DYNAMOS AND DYNAMO DESIGN 



59 



86. To realize either of the foregoing conditions in a 
series- wound dynamo, it is necessary to adopt some method 
of regulation, whereby either the effect of variations in the 
current on the magnetizing force of the field may be 
neutralized or the effective E. M. F. of the armature 
altered to suit the conditions. The former result may be 
obtained by placing an adjustable resistance in parallel with 
the magnetizing coil, as represented in Fig. 44. In this 
diagram, S.F. represents the magnetizing coil, or series- 
field, and R is the adjustable resistance, connected in 



8,F. 



r»' 




jjfO? 



/-I* 



±B 



,«z-« — - 




wwwvwwv 

Be 



PIO. 44 

parallel with the magnetizing coil, as described. It will be 
seen that the current divides between the two branches of 
this part of the circuit, and by varying the resistance R the 
proportion of the whole current that flows through the 
magnetizing coil S. F, may be varied as required. Other 
methods of varying the E. M. F. of series-wound dynamos 
intended for arc lighting will be described later. 

86. Series-winding is very little employed in dynamos, 
^xcept for machines designed to give a constant current, 
such as is used for operating lamps or other devices that are 
connected in series. For motors, however, series-winding 
is very useful, since when starting up under a heavy load, or 
whenever taking a current in excess of the normal amount, 



Digitized by VjOOQ IC 



60 



DYNAMOS AND DYNAMO DESIGN 



§13 



the field strength is increased, which iftcreases the amount 
of the reaction between the armature winding and the field, 
that is, increases the turning force of the armature. 

87. Characteristics of Series Machine. — In the ordi- 
nary series dynamo it is evident that the voltage at the 
terminals of the machine will vary greatly with the current 



























^^ — ' 




#,_ 




•-*"*** 










• 


^^- 












> 
/ 


f 




— ~ 


^ 


-j^ 






// 
1 1 
















11 














/ 








m — .- - 


J1lt*f- 






/ 
1 
u 






^ 


'0000(3 






ll 






1 [ 










•J 






E^£Mi 


d 




^ 


*^ 










■^ 












^ 













AMTfEnma i 

PXO. 45 



output, because the current flows around the field coils, and 
every change in the current is accompanied by a correspond- 
ing change in the E. M. F. In Fig. 45, the curve adc shows 



Digitized by VjOOQ IC 



§12 * DYNAMOS AND DYNAMO DESIGN 61 

the general shape of the external characteristic of a 
series- wound machine. If we draw the line O d represent- 
ing the volts drop in the armature and field and add 
its ordinates to those of the external curve, we obtain the 
total characteristic efg^ showing the relation between 
the load or current and the total voltage generated in 
the armature. 

It will be noticed that the general shape of this total char- 
acteristic curve is similar to an iron magnetization curve. 
As the current increases, the voltage rapidly increases until 
a point is reached where the iron begins to saturate. The 
curve then bends off to the right and the increase in volt- 
age becomes much smaller with the increase in current. 
Another point to be noted is that neither of the curves 
passes through point (9, because, even when no current is 
flowing, the machine generates a small E. M. F. O //, due to 
the residual field magnetism. When the external resistance 
is such that the machine is worked on the straight part of 
the curve, say between // and ;//, its operation will be very 
unstable. For example, if the current decreased a little, 
due to a slight increase in the external resistance, the field 
magnetization and also the voltage would decrease by quite 
a large amount ; this would cause a still further decrease 
in the current, the final result being that the machine would 
drop its load. If, however, the external resistance is low 
enough so that the current magnetizes the field beyond the 
bend of the curve, the action willbe stable because a reduc- 
tion in current is not then followed by a large reduction in 
the E. M. F. For every series machine there is, therefore, a 
certain resistance for the external circuit which, if exceeded, 
will cause the operation of the machine to become unstable. 
This resistance is sometimes called the critical resistance. 

In Fig. 45, it should be noticed that while the total char- 
acteristic keeps on rising slightly with the increasing cur- 
rent, yet the external characteristic drops after a certain 
point has been passed. This is because the increase in the 
armature drop more than offsets the slight increase in volt- 
age due to the additional field ampere-turns. 



Digitized by VjOOQ IC 



62 



DYNAMOS AND DYNAMO DESIGN 



§ia 



SnUNT WINDING 

88. The second method of self-excitation consists of 
forming a separate circuit of the magnetizing coils, which 
are connected directly between the brushes, or in shunt to 
the external circuit, this style of winding being, therefore, 
known as shunt winding:. This is illustrated in Fig. 46. 




PlO. 46 

It will be seen that the magnetizing-coil circuit is in a 
measure independent of the external circuit Re^ it being 
exposed at all times to the full difference of potential that 
exists between the brushes -\-B and —B ; from this it fol- 
lows that changes in the current flowing in the external 
circuit do not affect the magnetizing force acting on the 
field, except as they may change the difference of potential 
between the brushes. Changes in the current of the exter- 
nal circuit do affect this quantity in several ways; namely, 
by varying the drop due to the resistance of the armature 
winding, by varying the counter magnetomotive force of 
the armature winding, and by varying the length of the 
path of the lines of force by the variations in the amount by 
which they are distorted by the cross-magnetomotive force. 
This last is comparatively unimportant, but the other two 
require careful consideration in the design of dynamo 
machinery, as will be pointed out. 



Digitized by VjOOQ IC 



§12 DYNAMOS AND DYNAMO DESIGN 63 

89. In a shunt-wound motor the conditions are different, 
the magnetizing-coil circuit being supplied directly from the 
mains ; the magnetomotive force then depends simply on 
the difference of potential between the supply mains, which 
is usually kept constant, so that in general a shunt-wound 
motor may be considered as having a constant magnetizing 
force acting on its field magnet. 

90. Characteristics of Sliunt Maclilne. — In the case 
of a shunt dynamo there are two characteristic curves that 
are commonly drawn in order to indicate the performance 
of the machine. One of these is called the internal char- 
acteristic, and shows the relation between the voltage at 
the brushes and the current in the field, or, perhaps, what 
is more usual, the ampere-turns on the field. In order to 
obtain this curve, the machine is run at a constant speed 
without load and its field is excited from an outside source, 
so that the current in the shunt coils can be varied from a 
small amount up to or above the full current for which they 
are intended. The voltage obtained at the brushes will at 
first increase almost in direct proportion to the field current, 
but as the iron becomes saturated the increase in E. M. F. 
for a given increase of field ampere-turns will become less. 
The internal characteristic curve, therefore,' indicates at 
what current the field begins to saturate, and the general 
shape of the curve is the same as the dotted curve h efg^ 
Fig. 45. 

91. The external characteristic of a shunt dynamo is 
quite different from that of a series dynamo, because the 
field current is to a certain extent independent of the current 
in the external circuit. Fig. 47 shows the general shape of 
the external characteristic; the curve abcde shows the 
relation between the voltage at the brushes and the current 
in the main circuit. The total characteristic agh is found 
by adding the ordinates oi O k\.o those of abcde. The full 
line part abc oi the external characteristic represents the 
actual working range of the machine, assuming that 0/ 
represents the full-load current. At no-load, zero current, 



Digitized by VjOOQ IC 



64 



DYNAMOS AND DYNAMO DESIGN 



§ia 



the voltage at the brushes is represented by the vertical O a. 
This is the maximum voltage that the machine is caf)able of 
giving with a given field excitation, because the pressure 
across the field coils is a maximum; as soon as current is 
taken from the armature, the pressure across the fields 
decreases and there is also a drop in the armature, hence the 
voltage decreases. It is evident that a shunt machine can 
generate its full voltage even if the external circuit is open 















































• 






























~^ 








f^^ 


"^^■i 








a 


mn$ r\ 


«M 








**" 


^^"^^ 


V 

X 







)W0"0( 









L- 


>©1 


-J 








*-^ 


• 


' 


I 






n 




•*' 


^^-»'' 








1 

• 
1 - 






^# 


LjJ^ 




r_^ 1 


^^^..^ 


— ' — 




1 
1 

1 


o 








XICJ» 


&BMBJ 


r 


J 






% 



Fio. 47 

because the path through the magnetizing coils is not inter- 
rupted as it is with the series dynamo. As the current is 
increased from no load to full load, the terminal voltage 
drops from Oa to fc. In this respect the shunt dynamo 
behaves somewhat like a separately excited machine, but 
the falling off in voltage with a given machine would be 
greater with shunt excitation than with separate excitation. 
The reason for this is, of course, that with the separately 
excited machine the pressure applied to the field coils is 
constant, whereas, with the shunt dynamo every falling 



Digitized by VjOOQ IC 



§ 12 DYNAMOS AND DYNAMO DESIGN 65 

off in terminal pressure causes a decrease in the field 
excitation. 

92. If the external resistance is made low enough, the 
current will reach such a value, represented by (?/, that the 
pressure across the shunt terminals becomes so low that 
the unstable portion of the magnetization curve is reached. 
When this is the case, any further decrease in terminal 
voltage causes the machine to drop its voltage entirely, the 
unstable portion of the characteristic curve being repre- 
sented by the dot-and-dash part de. In well-designed shunt 
machines, this unstable condition is not usually reached 
until the current is considerably greater than the full-load 
rating of the machine. For example, in the case shown in 
Fig. 47 the working range of current is from O to/, and the 
action would not become unstable until the current increased 
to /. A peculiarity of the shunt machine is, therefore, that 
the external resistance must be above a certain critical 
value in order for the machine to generate. For example, 
a shunt machine, if short-circuited, will drop its voltage. 
This is just the opposite to a series machine, where the 
external resistance must be below a certain critical value in 
order to allow the machine to pick up its voltage, and in 
case of a short circuit on a series dynamo, the voltage rises 
very rapidly. 

93. If the armature of a shunt dynamo has a very low 
resistance and the field a high resistance, the curve abc will 
drop but little, and the machine will, therefore, hold its 
voltage fairly constant within the working range. The 
smaller the slant oiabc^ the greater must be the external 
current before the voltage becomes unstable. The falling 
off in voltage can be compensated for by cutting out some 
of the resistance in the shunt field, so that by resorting to 
the rheostat the line abc can be made horizontal or even 
rise a certain amount as the load comes on. This method of 
regulation is feasible when the load does not vary suddenly, 
but where a constant potential is desired it is now customary 
to use compound-wound machines, 

44 — 6 



Digitized by VjOOQ IC 



66 



DYNAMOS AND DYNAMO DESIGN §12 



COMPOITNT) TTRTDING 

94. From the foregoing statements it will be seen that 
in order to maintain a constant difference of potential 
between the brushes of a dynamo (assuming a constant 
speed), the magnetomotive force of the magnetizing coils 
must be increased as the current increases, both to increase 
the number of lines of force so as to increase the E. M. F. 
generated, and to make up for the counter magnetomotive 
force of the armature winding. One way to accomplish 
this result is to place an adjustable resistance r, Fig. 46, in 
the magnetizing-coil circuit, which may be gradually cut out 
as the current output increases, thus reducing the resistance 

of the magnetizing- 
coil circuit, and in- 
creasing thereby the 
current flowing 
through it. This, 
however, requires per- 
sonal attention, and 
in case the current 
from the dynamo fluc- 
tuates rapidly, it is 
difficult to operate the 
resistance with suffi- 
cient rapidity. Since 
the amount by which 
the magnetomotive force of the magnetizing coils must be 
varied is closely proportional to the current flowing, which 
follows from the nature of the causes that require the 
variation, it is possible to obtain the required variation by 
providing additional magnetizing coils through which the 
main current passes. This is known as compound Mend- 
ing, and is illustrated in Fig. 48. 

96. It is evident that this is a combination of series and 
shunt winding, the shunt winding furnishing an approxi- 
mately constant magnetizing force and the series-winding 
an additional magnetizing force that is proportional to the 




Digitized by VjOOQ IC 



g 12 DYNAMOS AND DYNAMO DESIGN 67 

current output of the machine. This latter winding is so 
proportioned that it furnishes the proper increase in the 
magnetomotive force, as the current increases, to make up 
for the dropping off of the difference of potential between 
the brushes that would otherwise occur. For certain classes 
of work, a little more than this amount is provided, so that 
the difference of potential between the brushes rises slightly 
as the current output increases. In such a case the machine 
is said to be over-eoin pounded. Compound-wound ma- 
chines are provided with a rheostat in the shunt field, but 
this rheostat is not intended for regulating the vokage in the 
sense that it is used with a plain shunt machine. The rheo- 
stat is intended to allow an initial adjustment of the voltage, 
thus compensating for any departure from the standard speed 
or for variation in the quality of the iron in the magnet frame. 
Also the shunt winding heats up after the machine has been 
in operation for a while and this heating raises the resist- 
ance, thereby cutting down the magnetizing current. If a 
rheostat is included in the field circuit this can be compen- 
sated for by cutting out some of the resistance after the 
machine has become warmed up. The series-winding, how- 
ever, compensates for all changes in the voltage due to 
changes in the load. 

96. Compound winding is not used nearly so often for 
motors, since either a series or a shunt winding serves for 
almost all conditions of operation. Nevertheless, for appli- 
cation to such machinery as printing presses, a compound 
winding is extremely useful, as the series-turns produce a 
powerful field at starting and at slow speed, and they may 
gradually be cut out or connected in various combinations 
to produce different working speeds without the necessity 
of inserting an external resistance in the armature circuit, 
except for starting up, when a resistance may be tempo- 
rarily used. 

97. Characteristics of Compound-Wound Machines. 

The characteristic of a compound-wound dynamo will depend 
very largely upon the magnetizing effect of the series-coils. 



Digitized by VjOOQ IC 



68 



DYNAMOS AND DYNAMO DESIGN 



12 



The coils may be strong enough to overcom pound the 
machine, in which case the external characteristic would 
take the form of a rising line ade^ Fig. 49. If the series- 
coils were just strong enough to keep the terminal voltage 
constant, the characteristic would become a horizontal 
line ac. If the series-coils were not powerful enough to 
compensate for the falling off in voltage, the characteristic 
would drop as shown by af^ but the dropping would not be 



i 













































.■ 










•- 




• 




















/- 














ah^n% 












0"6 


ffw^ 


V) 








L_ 


H 


^000 

*) — 


\ 
















^ 




- 
















XecMl 













AMBBBBSM 

Pig. 40 



as great as for a plain shunt machine. Usually compound- 
wound machines are wound to give a rising characteristic 
so that if the external resistance is lowered very greatly, as 
by a short circuit, the effect is to put a very heavy over- 
load on the machine because the voltage will not drop on 
short circuit as with a shunt dynamo. In connecting up the 
series-coils of a compound-wound machine, care must be 
taken to see that they are connected so as to aid the shunt 
coils and not oppose them, because in the latter case the 
machine would drop its voltage as soon as a load came on. 



Digitized by VjOOQ IC 



§ 12 DYNAMOS AND DYNAMO DESIGN 69 



BUHiDra'G "DP THE FIELiD 

98. Any iron, after being magnetized, retains a certain 
amount of residual magnetism, so that there will be a small 
E. M. F. generated in the armature winding when the 
armature is rotated and the field circuit left open; this is 
utilized to start the current in the magnetizing coils. In 
the case of a shunt-wound dynamo, when the machine is 
started and the magnetizing-coil circuit closed, the small 
E. M. F. generated in the armature by the residual magnet- 
ism sends a small current through the magnetizing coils, 
producing a small magnetizing force. If this magnetizing 
force tends to send lines of force through the magnetic cir- 
cuit in the same direction as the residual magnetism, the 
number of lines of force will be increased; this will increase 
the E. M. F., which increases the current in the magnet- 
izing coils, and still further increases the number of lines of 
force and the E. M. F., which process continues until fur- 
ther increase in the magnetizing force results in so little 
increase in the number of lines of force that the E. M. F. 
generated becomes steady, the windings being so designed 
that this shall be the E. M. F. at which it is desired to run 
the machine. 

It will be seen that if the external circuit is open, all 
the current that the E. M. F. (due to the residual magnet- 
ism) produces flows through the magnetizing coils ; if, how- 
ever, the external circuit is closed, only a part of the current 
flows through the magnetizing coils, so that the field will 
build up more slowly than with the external circuit open, 
and, in fact, will not build up at all if the external resistance 
is low, as compared with the armature resistance. From 
this it follows that a shunt-wound machine should* be started 
up with its external circuit open. 

A series-wound machine, on the contrary, must have its 
external circuit closed in order that any current may flow 
through the magnetizing coils, and the lower the resistance 
of the external circuit, the more quickly will the machine 
build up. 



Digitized by VjOOQIC 



70 DYNAMOS AND DYNAMO DESIGN § 12 

From the 'above it will be seen that a com pound -wound 
dynamo may be started with its external circuit either open 
or closed, since it has both series and shunt wound coils. 
Usually, however, such machines are started and brought 
to their full E. M. F. with the external circuit open. 

99. At starting, while the current is increasing in the 
magnetizing-coil circuit, the self-induction of the magneti- 
zing coils increases their apparent resistance, and a part of 
the energy supplied to the coils is stored up in the magnetic 
field that is being established. As soon as the current in 
the magnetizing coils reaches its maximum value, how- 
ever, and so long as it remains constant at this value, 
the entire amount of energy delivered to the coils is 
expended in heating the wire ; that is, it requires (directly) 
no energy to maintain a magnetic field at a constant 
value, the field depending on the number of ampere-turns 
that are acting on the magnetic circuit. It is obvious, 
however, that in order to force the current through the 
wire of which the magnetizing coil is composed, energy 
must be expended, but this energy appears entirely as heat, 
and, consequently, is wasted as far as any practical applica- 
tion of it is concerned. The number of watts expended in 
sending the current through the magnetizing coils should, 
therefore, be made as small as the design of the machine 
will permit, both to prevent any excessive waste of energy 
and to prevent possible damage by the heat liberated. In 
practice, the loss of energy from this cause varies from 
about 1 per cent, of the total output of the machine in 
larger sizes, to 5 per cent, or more in the smaller. 

100. In shunt- wound machines the magnetizing coils 
are exposed to the full difference of potential that exists 
between the brushes of the machine, and, consequently, 
should use only a small amount of current in order that the 
loss in watts may be the required small percentage of the 
output. From this it follows that the wire used for the 
magnetizing coils should be of small size and of considerable 



Digitized by VjOOQIC • 



§l5i DYNAMOS AND DYNAMO DESIGN 71 

length, making a large number of turns around the mag- 
nets, both to give the necessary resistance to keep the cur- 
rent at its proper value and to allow of this small current 
furnishing the requisite number of ampere-turns. In series- 
wound machines, however, as the total current flowing 
gives the magnetizing force, the magnetizing coils need to 
have comparatively few turns, which should be of corre- 
spondingly large wire, in order that the watts loss (equal to 
PR) should be kept within the desired limits. 

It will be seen that in series-wound dynamos the differ- 
ence of potential between the terminals of the machine is 
less than that which appears between the brushes by the 
amount of the drop in the magnetizing coils. 

The above remarks concerning the magnetizing coils of 
shunt and series wound dynamos also apply to those of 
compound-wound machines, since they are made up of a 
shunt-winding and a series-winding. 



Digitized by VjOOQIC 



Digitized by VjOOQ IC 



DYNAMOS AND DYNAMO DESIGN 

(PART 3) 



DIAGRAMS OF CLOSED-COIL WESDINGS 



RING WINDINGS 

1. Iting T^ndlnfiTS are not much used in modern ma- 
chines, although the single multicircuit ring winding has 
advantages that make it quite promment for generators 
producing high voltages on one hand, or very large currents 
on the other. This winding, arranged in a four-pole field 
magnet, is shown in Fig. 1. It will be noticed that there 
are two turns per coil, and that the winding is represented 
as continuous, leads being tapped in for connecting to the 
commutator. Such construction is exactly equivalent, elec- 
trically, to that where the ends of each coil are brought 
down to the commutator. The latter method, however, is 
preferable to the former, because the connections are more 
easily made at the commutator. 

2. To connect the machine to the outside circuit, the 
two positive brushes should be connected together to form 
the positive terminal, and the two negative brushes in the 
same manner to form the negative terminal. The direc- 
tion of flow of the armature currents is shown in the four 

§13 

For notice of copyright, see page immediately following the title page. 



Digitized by VjOOQ IC 



a DYNAMOS AND DYNAMO DESIGN §13 

paths. It will be observed that in each neutral region a 
single armature coil is short-circuited by a brush, and 
therefore has no current in it, as indicated by the absence 
of the arrowheads. Under the north poles it is found 
that the arrowheads point inwards, while under the south 
poles they point outwards ; of course these conditions would 
be exactly interchanged were the direction of rotation 
reversed. 

3. Comparing Fig. 12 of Part 1 with Fig. 1, it will be 
noticed that they are very similar as far as the winding of 

the coils is concerned, 
^ the only difference 
^ being in the number of 
^ coils on the complete 
^ armature. Were the 
armature of Fig. 1 
put in a six-pole or an 
eight-pole field mag- 
net, it would not be 
necessary to alter any- 
^ thing connected with 
^ the winding in the 
^ least degree, but the 
^ brushes, of course, 
^ would have to be the 
same in number as the 
poles. It follows that 
an armature provided with this winding may be used in a 
field frame with any number of poles. While this is an 
interesting feature possessed by no other winding but the 
multicircuit ring type, it is of little commercial importance, 
as it is very unusual to have occasion to use an armature in 
several machines having different numbers of poles. 

4. In Pig. 1 there are thirty-two coils on the armature, 
or eight between brushes of opposite signs. Suppose that 
these coils were wound to generate 100 volts each ; then, since 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 8 

there are seven active ones between brushes, the machine 
would deliver about 700 volts. It will be observed that the 
coils connected to one set of brushes are separated by a 
considerable distance on the armature from those connected 
to the other set. In fact, the extreme E. M. F. between 
one coil and its neighbors can be only that developed by the 
one coil. To insulate such a coil from others does not 
require very thick insulation, and with each coil insulated, 
the complete winding will be successfully insulated. Thus, 
to insulate each coil to withstand 100 volts between it and 
its neighbor, will successfully insulate the complete arma- 
ture for a pressure of 700 volts. This is a great advantage 
that ring windings with a single coil in series between 
segments have over drum types, for in the latter the coils 
overlap one another, and the room required for insulation, 
were these windings used for several thousand volts, would 
make the complete armature bulky and cumbersome. On 
account of the ease with which they can be insulated for 
high pressures, ring windings are much used for series arc- 
lighting dynamos. 

5. In machines intended for generating very large cur- 
rents, as, for instance, large, direct-current, 125-volt gen- 
erators, it is found necessary, in order to prevent sparking, 
to use as many commutator segments as possible. The 
greatest number of segments possible is one segment per 
turn ; on a ring armature a turn includes but one face con- 
ductor, while for a drum armature it includes two face coh- 
ductors. 

It will be shown later that the voltage depends on the 
number of face conductors in series ; so with the same volt- 
age, i. e., the same face conductors, a ring winding may 
have twice as many segments as can be used for the drum 
type. 

Ring windings have a further advantage that, since the 
coils do not overlap one another, any coil or coils may be 
removed without disturbing any others, and reinsulated or 
replaced by new ones, if necessary. 



Digitized by VjOOQIC 



DYNAMOS AND DYNAMO DESIGN §13 



DRUM WINBINGS 



PABAI^USL WINDINGS 

6. Diagrams of dram windingrs are far more compli- 
cated than those for ring windings, and it is necessary to 
adopt some conventional way of representing them. The 
method that will hereafter be used is shown in Fig. 2, which 



Pig. 2 

explains itself. The face conductors are here shown as short 
radial lines outside of which are the rear end connections; the 
position of the poles is indicated as shown. Inside of the 
radial lines are shown the front end connections and leads to 
the commutator, inside of which are the commutator and 
the brushes. As inferred above, the commutator end of an 
armature is called the front end. 

Almost all modern electric generators of 50 kilowatts, or 
more, output have but a single turn to a coil, and in many 



Digitized by VjOOQ IC 



Digitized by VjOOQIC 




Pig. I 



Digitized by VjOOQ IC 




Crff55 ^ecf/an of S/af% 



^d9 



(b) 




C/VS5 5ect/o/t tf^/^^ 




Digitized by VjOOQIC 



Digitized by VjOOQIC 



§13 DYNAMOS AND DYNAMO DESIGN 6 

cases, instead of using round copper wires, rectangular bars 
are used and the windings thus made are termed bar- 
-wound, to distinguish them from those that are Tvlre- 
%\ound. Where the output of a generator is made the 
greatest possible, in order to be economical of material, the 
available room for the windings scarcely permits the use of 
round wires on account of the waste space left between wires, 
and the bar winding must then be used. Bar windings are 
sometimes used with coils of two or more turns each, but 
very rarely, as the coils with more than one turn are hard to 
handle; when made of large copper conductors, and besides 
it is seldom that a machine requiring large bars for con- 
ductors also requires a two-turn winding. 

7. Fig. 3 is a diagram of a four-pole, single, parallel, 
drum winding with seventeen coils, seventeen commutator 
segments, and thirty-four slots. The complete diagram is 
shown at (a), while the details of a slot and a sketch of a 
wire-wound coil are shown at {6) and the same for a bar- 
wound coil at (c). It will be noticed that the current comes 
in through brushes Ay A' and goes to segments a, 6, and o. 
To a is connected one lead, or terminal, of coil 1-10, the 
other end of which is connected to segment 6. This coil is 
short-circuited by the brush A and has no current in it 
at the instant shown. The face-conductors of coil 1-10 
are about midway between the poles or in the neutral 
region. From a is connected another coil 8-33, which ter- 
minates at 7, where, also, connects coil 6-^31, etc. Thus 
one of the paths is a,S-33 to y, 6-31 to x, 4-29 to v, 2-27 
to /^ where it touches positive brush B, In the same way 
another path is from b to coil 3-12 to c, 6-H to d, 7-16 
to e, 9-18 to /, and thence to positive brush B', There is 
also a path that includes the segments from brush A' to B 
and a path including segments from A' to B'. The winding 
is seen to have four paths or as many as there are poles, and 
there are four coils in series in each path. The face con- 
ductors carry the current away from the commutator under 
north poles and toward the commutator under south poles. 



Digitized by VjOOQ IC 



6 DYNAMOS AND DYNAMO DESIGN §13 

This is the reverse of the winding shown in Fig. 1, the rea- 
son being that the direction of rotation is reversed. The 
paths and the direction of currents in the windings should 
be carefully studied until thoroughly understood. 

8. In following out the various paths, it will be observed 
that the end connections of the coils on the commutator end 
are not essential to determining a path, and since they con- 
fuse the diagram, it is usual to omit them. This makes the 
diagram appear as though the coils were of a single turn 
only, and in the diagrams following they will be drawn this 
way, it being understood that the diagram of the winding is 
exactly the same, regardless of the number of turns per coil. 

It is seen from Fig. 3 that the position the brushes occupy 
on the commutator with regard to the pole pieces depends 
on the way in which the terminals of the coils are connected. 
For example, take the lead 1-a that connects from slot 1 to 
bar a. If this lead were brought out to the bar directly in 
front of the slot, then the brushes A, B, A\ B' would, when 
in the neutral position, be opposite the center of the space 
between the poles. This would, howev,er, necessitate one 
long terminal and one short terminal on each coil, because 
the terminal lO-d would have to be long enough to reach 
around to the bar adjacent to that opposite slot i. It is 
usual, therefore, to bring the coil leads, when connecting 
them to the commutator, around through an angle of about 
one-half the pitch of the poles. For example, in Fig. 3, the 
lead 1-a is brought around to the right through an angle of 
about one-eighth a circumference, thus bringing the brushes 
opposite the centers of the pole pieces when in the neutral 
position. This arrangement makes the coil terminals i-ii 
and 10-d of about equal length and renders the connections 
symmetrical. Bringing the coil terminals around in thia 
way is termed giving them a lead, and it is evident that the 
neutral position of the brushes will depend on the lead. For 
some kinds of railway motors, one coil terminal is brought 
out straight, so that each coil has one long terminal and one 
short one; but with nearly all generators the coil terminals 



Digitized by VjOOQ IC 



Digitized by VjOOQIC 




Pia4 



Digitized by VjOOQIC 



SIcT' ^ j L ^^ 




Digitized by VjOOQIC 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 7 

are brought around through an angle of about one-half the 
pole pitch, and the neutral position of the brushes at no load 
is nearly opposite the center of the pole pieces. 

9. At (*), Fig. 3, is shown the details of a wire-wound 
coil and the cross-section of a slot. The slot is first lined 
with insulating material, then the coil is put in place, and 
lastly a piece of wood is usually inserted over the coil to 
protect it. Coils 1-10 and 3-12, the ones that are shown in 
heavy lines at {a), are shown inserted in the slots in the core 
with their leads properly connected to the commutator seg- 
ments a, b, c. At {c) the same slots are shown, each occupied 
by a single, rectangular, copper bar instead of a number of 
wires, as at {b). The same coils are shown connected to the 
segments a, b, c. Comparing {b) and (r), it will be noticed 
that the latter is much simpler in appearance, for the coils, 

' consisting of but a single turn, have no end connections on 
the front, or commutator, end, as do the coils {b), 

10. Suppose that all the coils were connected to the 
commutator thus: connect 1 to segment b and 10 to seg- 
ment a, 3 to segment c and 1£ to segment by etc. ; in other 
words, if the leads of each coil were interchanged, what effect 
would it have on the action of the winding ? It would still 
be a single parallel winding, since the coils terminate in 
adjacent segments and the E. M. F. generated in the coils 
would be the same in direction and amount; but since each 
coil has had its leads reversed, the E. M. F. of the com- 
pleted armature would be reversed and the brushes A and A' 
would become positive brushes, while B and B' would become 
negative brushes. Each lead would have to be a little longer 
than before and, therefore, more copper would be required ; 
such a winding would therefore be objectionable on this 
account. 

11. It will be noticed in Fig. 3 that at either end of the 
core the coils extend both to the right and to the. left. As 
has been explained, the cross-connections interfere with one 
another unless involute end connections are used, and in 
order to avoid this it is usual to make the winding in two 



Digitized by VjOOQ IC 



8 DYNAMOS AND DYNAMO DESIGN §48 

layers,. placing two coils in each slot, one on top of the other, 
so that all those in the bottom extend to the right, say, 
while those at the top extend to the left. Such windings 
are called two-layer windingps, while that of Fig. 3 is called 
a slngrle-layep winding:. 

In Fig. 4 is shown a two-layer winding of one turn per 
coil, the exact electrical equivalent of Fig. 3, but having only 
seventeen slots instead of thirty-four. 

12. In Figs. 3 and 4 notice that at the instant shown 
coil 1-10 ends in segments connected to the negative brush ^, 
hence is at the potential of A ; likewise for the same reason 
coil 27-2 is at the potential of positive brush B^ so that the 
potential difference between these two coils is equal to that 
developed by the machine at the instant shown. Conduc- 
tors 1 and 2^ between which there exists this large poten- 
tial difference, are adjacent in both diagrams, being in the 
same slot in Fig. 4. It will be noticed that this is a neces- 
sary property of drum windings and may be stated thus: 
Between adjacent coils of a drum-armature winding at times 
during a revolution there exists the full potential of the 
machine. This constitutes the chief objection to drum 
windings, and necessitates that extra-good insulation be 
provided between the coils in the slots [see {b). Fig. 4]. 
Observe that the extreme difference of potential also exists 
between conductors 33 and 31^, 11 and 12, 19 and 20, and 27 
and 28. In general, it may be said that the extreme differ- 
ences of potential will exist between one coil of one layer and 
its neighbor in the other layer at the time of commutation. 

13. In Fig. 4, the paths or circuits from negative to 
positive brushes are as follows: 

A, 3-12, 5-1 U, 7-16, 9-18, B' 
A, 8-33, 6-31, 4-29, 2-27, B 
A', 19-28, 21-30, 23-32, 25-3 4, B 
A\ 26-17, 2^-15, 22-13, 20-11, B' 

This includes all coils but 1-10, which at the instant shown 
is short-circuited. 



Digitized by VjOOQIC 



Digitized by VjOOQIC 




PlO. 5 



Digitized by VjOOQIC 



Sar. 



Arr^rtfement tf^SM;, 








Digitized by VjOOQIC 



Digitized by VjOOQIC 



§13 DYNAMOS AND DYNAMO DESIGN 



14# A diagram of a single series- windlngf of the single- 
layer type is shown in Fig. 5. This winding is exactly like 
that of Fig. 3, except in regard to the connections to the com- 
mutator, and both are open to the same objection that being 
in a single layer the end connections will interfere if the 
barrel type of winding is used. In these two diagrams, 
comparing (r), it will be noticed that in Fig. 3 the winding 
of one coil laps over that preceding it, while in Fig. 5, the 
winding progresses in a series of zigzag lines or waves. The 
former winding is therefore sometimes called a lap ipvlnd- 
Ingr and the latter a wave Tvindlngr, but the terms parallel 
and series are to be preferred, since they are applicable to 
both ring and drum types of windings, while the terms lap 
and wave are only applicable to the drum type. 

15. The paths through the winding, Fig. 5, are as 
follows : 

_ {A, 1^10, 19-28, 3-12, 21-30, 6-U, 23-32, 7-16, 25-3 i, 5 ^ 
•( A, 26-17, 8-33, 2^-15, 6-31, 22-13, jh29, 20-11, B)^ 

At the instant shown, there are eight coils in one path 
and seven in the other, there being a series of two coils 9-18 
and 27-2 short-circuited by the brush B. Since there are 
seven coils in one path and eight in the other, it would appear 
that the winding is not symmetrical and one path would take 
more current than the other, but turn the armature a small 
fraction of a revolution, and it will be found that the path 
that now has eight coils may then have seven and that with 
seven may then have eight, so that the two paths will have 
the same average number of coils, even though they may 
not be alike at a given instant. In actual windings, there 
are usually from twenty to fifty or more coils per path, 
and the differences caused by having one coil more or less 
in a path where the number is so great is insignificant. 

16, Comparing the number of coils in series in each 
path of the windings. Figs. 3 and 5, shows that the latter 

44—7 



Digitized by VjOOQIC 



10 DYNAMOS AND DYNAMO DESIGN §13 

has twice as many as the former and, if the coils are the 
same in both cases and only the commutator leads changed, 
the latter will give twice the voltage of the former with but 
half of the Current capacity, since it has but half the paths. 
The watts output of the winding remains unchanged, how- 
ever, since doubling the number of volts and halving the 
current will give the same number of watts. 

With the series-windings, but a single pair of brushes is 
required to collect the current, as shown, but the other 
neutral points may be also supplied with them, as shown by 
the dotted lines. 

n. The winding fulfils the formula that the number of 
coils 

<r = |F±l = |x9-l = 17, 

nine being the span of the coils on the commutator seg* 
ments, that is, from a to o; the terminals of coil 1-10 are 
nine segments removed from one another. It will be seen 
that this same number of coils can be made into another 
winding with a pitch of eight, thus, 

c = |f±i = 1x8 + 1 = 17 

This winding is shown in diagram in Fig. 6, which is 
made after the two-layer type. Figs. 6 and 6 are electrically 

the same, except that in the former a series of ^ coils 

spanned a circumference plus one segment on the com- 
mutator, while in the latter the series lacks one segment of 
spanning a circumference. The effect of this is to inter- 
change the connections of the series of coils, so that the 
brushes exchange signs, although the direction of current in 
the face conductors is as before. It has the same effect as 
though the terminals of each coil were reversed at the com- 
mutator, as was explained for the parallel type of winding. 
To actually reverse the leads of each coil of the series-type 
of winding would require such extra-long leads that it would 
be impracticable. 



Digitized by VjOOQ IC 



Digitized by VjOOQIC 




Fli 



Digitized by VjOOQIC 




Digitized by VjOOQIC 



Digitized by VjOOQIC 



§13 DYNAMOS AND DYNAMO DESIGN 11 

18. Where drum-armature windings consist of but a 
single turn per coil, and often where there are many turns 
per coil, an inspection of the armature will usually show 
whether it is of the parallel or series-type in the following 
way: Consider face conductor i. Fig. 4, which is of the 
parallel type, and note that both of the end connections 
extend in the same direction. The same is true of con- 
ductors ^, 5, 7, etc., as well as the even-numbered conductors, 
which, shown dotted, are underneath and are not visible on 
the completed armature. Referring to Fig. 6, it will be seen 
that the end connections on the rear end extend to the right 
from conductor i, while on the front end they extend to the 
left. It follows that if the end connections from a face con- 
ductor extend the same way on both ends, the armature 
winding is of the parallel type; while, if the end connec- 
tions do not extend the same way, the winding is of the 
series-type. Whether the winding is single or double can- 
not usually be determined from the completed armature, 
except by testing the resistance between segments. 

It will also be noticed that in Figs. 4 and 6 the coils 
span four slots on the armature, that is to say, coil 1-10 lies 
in slots 1 and 6. With seventeen slots and four poles, four 
and one-quarter slots is the exact pole pitch and four is 
the nearest whole number. 

19. Double Parallel Winding. — A double parallel 
TTlndingr could be readily made with seventeen coils and 
four poles by connecting coil 1-10 to segments a and r, coil 
S-12 to segments b and d, etc., but such a diagram is not 
very instructive, for the brushes must be made thick enough 
to span about one and one-half segments at least, and with 
so few segments in the commutator, there would be but two 
coils in series per path, which does not approach practical 
requirements at all. However, the double parallel winding 
is so simple that it should be understood from what has been 
said, without the aid of a special diagram. 

20. I>ouble Series- Winding. — A double serles-Tvind- 
ing, however, cannot be made from seventeen coils and four 



Digitized by VjOOQIC 



12 DYNAMOS AND DYNAMO DESIGN §18 

poles, because the number of coils must fulfil the formula, 

C = ~ F ± 2 ; and since ^ in this case is even, C should 

be even regardless of whether Y is even or odd. Just why 
seventeen coils will not answer is seen by remembering that 

in a double series-winding, a series of^ coils should encircle 

the commutator with two segments over or less, or in this 
case two coils should span either fifteen or nineteen seg- 
ments. All coils must be alike, so since fifteen and nineteen 
are not divisible by two, the winding is impossible. 

In Fig. 7 is shown a double series-winding for a six-pole 
armature with twenty-eight segments and coils and twenty- 
eight slots. The number of coils in this winding fulfils the 

formula C = 4f±2 = -X 10 — 2 = 28, and since the 

pitch Y is divisible by two, the winding is doubly reen- 
trant. The four paths are as follows, the numbers referring 
to the commutator segments to which the coils are attached; 



e-ie, 16-26, 26-8, 8-18, 18-28, 28-10 
5-23, 23-13, 13-3, 3-21, 21-11, 11-1 
15-26, 25-7, 7-17, 17-27, 27-9, 9-19 
24-ljt, U-^, 4-22, 22-12, 12-2, 2-20 



+ 



Coils 6-24, 1-19, 5-15, and 10-20 are short-circuited by 
brushes at the instant shown and are not active. It will 
be noticed that coils are not short-circuited by a single 
brush, as in the case of parallel windings, but are short-cir- 
cuited by two brushes and the connections between brushes 
of like sign. The coils lie in slots 1 and 6, 2 and 7, etc., 
nence, the span of the coils on the armature is fivt^ slots, 
which is the nearest number to twenty-eight divided by six, 
the number of poles. 

21. Bucking^. — Many differerit windings have been pro- 
posed and used successfully, but the majority of armatures 
are wound with either the single parallel ring winding, 
the single parallel drum winding, or the single series- 
drum winding. The single parallel drum is by far the most 



Digitized by VjOOQ IC 



Digitized by VjOOQIC 




Digitized by VjOOQ IC 




Digitized by VjOOQIC 



Digitized by VjOOQIC 



§13 DYNAMOS AND DYNAMO DESIGN 13 

popular. For machines with six or more poles, with this 
winding, it becomes difficult to get the currents in the 
various paths to divide properly, especially if the poles for 
any cause should become of unequal strength. Suppose, for 
example, that the armature, through wear of the bearings, 
approaches appreciably closer to one pole than to the others. 
Since the air gap at this pole is less than the others, with 
the same magnetizing coils, this pole will become stronger, 
and the pole opposite it weaker; hence, the armature will 
be drawn towards the stronger pole by the unbalanced 
magnetic pull, which therefore tends to make the difficulty 
worse. Now, the conductors on one side of the armature 
will be in a stronger field and will therefore develop a 
greater E. M. F. than those on the other side, and the path 
that develops the greatest E. M. F. tends to take the great- 
est share of the current. If the E. M. F.'s developed by 
the various paths are very different, so that the E. M. F. of 
one path may overpower that of some other, heavy local 
currents will be developed within the armature. These are 
usually accompanied by trembling and groaning of the 
machine, due to excessive mechanical strains, and more or 
less violent sparking or flashing at the brushes. This phe- 
nomenon is called bucking. On account of the effects of 
armature reactions, it is somewhat more liable to occur in 
motors than in generators. 

22. Cross-Connectloii of Armatures. — To prevent 
bucking, and to equalize the division of the currents in par- 
allel-wound armatures, they are often crosB-conn^cted. 
It has been stated that points on the commutator that are 
two poles apart always remain at the same potential as the 
armature revolves. By cross-connection is meant the per- 
manent connecting of points in the windings that always 
remain at the same potential. This is done by means of 
copper rings, Fig. 8 (*), which are connected to the winding 
at intervals exactly two poles apart. In the figure, the rings, 
seven in number, are attached to the rear end, and the 
leads S, S from the winding can be plainly seen. Each ring, it 
will be noticed, is connected to seven points in the winding; 



Digitized by VjOOQ IC 



U DYNAMOS AND DYNAMO DESIGN §13 



00 



Digitized by VjOOQIC 



§ 13 DYNAMOS AND DYNAMO DESIGN 15 

hence, the armature is intended for a field magnet with 
seven pairs of poles, or fourteen poles. The armature shows 
the method of construction followed by the General Electric 
Company. Many other builders put the cross-connecting 
rings beneath the winding on the commutator end, making 
the connections to the commutator segments in the same 
manner as for the coils. In order that a winding shall prop- 
erly cross-connect, it is necessary that there be a whole 
number of segments for every two poles ; or, in other words, 
the number of segments should be divisible by the number 
of pairs of poles. In Fig. 8 (a), the armature bars are seen 
at ^, a; i^ i are the leads connecting to the commutator A, 
Double windings, both parallel and series, are not very 
reliable, and are but little used. They are more liable to 
sparking than are single windings, and if the commutator 
becomes in the least irregular, a segment higher than its 
neighbor will raise the brushes and open the circuit to the 
winding with the lower segment. A destructive spark results, 
and the low segment will be burned lower, thus aggravating 
the trouble. 



OPEN-COIL ARMATURE WINDINGS 

23. In Figs. 7 and 10 of Part 1, are shown typical open- 
coll armatures, one with the drum type and the other with 
the ring type of coils. It will be observed from the manner 
of connecting these coils to the commutator that they do 
not form closed circuits, for in them only one terminal, or 
lead, of a coil connects to a commutator segment; hence 
the name open-coil. These windings are now used commer- 
cially in America on three types of dynamos. The Brush 
and T-H arc dynamos, made by the General Electric Com- 
pany, and the Westinghouse arc dynamo, all of which are con- 
stant-current machines; that is, they are machines that are 
provided with some means of regulating the current so as to 
keep it at some fixed value, usually either 6.8 or 9.6 amperes. 
They are all used for series-arc lighting, the arc lamps having 
been found to operate well at the currents given. 



Digitized by VjOOQ IC 



16 DYNAMOS AND DYNAMO DESIGN §13 

24. Consider the action of the armature shown in Fig. 10 
of Part 1 ; the current entering by the negative brush finds 
but a single path to the positive through one of the sets of 
coils, and as the armature rotates, the brushes slide from 
one pair of segments to the other, and the inactive coil now 
becomes active, carrying the whole current, and so on. 
Since there is only one set of coils active at once, it is evi- 
dent that for economy such armatures should not have 
many coils, as their period of activity would be shortened. 
With brushes and commutator as shown, the current of the 
machine will be very suddenly shifted from one path to the 
other, and this will cause serious sparking on account of 
the self-induced E. M. F. of the coil. A conductor carrying 
current tends to surround itself with a magnetic field, and a 
coil, especially one surrounding iron, will become a powerful 
magnet, as has already been explained. If this current is 
varied, the magnetic field must change in proportion; and if 
reversed, the field must also be reversed. The changing of 
the number of magnetic lines of force through a loop or coil 
is the same in effect as the cutting of lines of force with its 
conductors, and induces an E. M. F. in the coil independ- 
ently of any cutting of the magnetic field emanating from 
the poles. Thus, the E. M. F. of self-induction is propor- 
tional to the rate of change of self-induced magnetism ; or 
since this varies directly with the current, it is proportional 
to the rate of change of current. The direction of this self- 
induced E. M. F. is such as to tend to prevent the increase 
of current and to tend to maintain it after the current has 
been established; it causes the currents to act as though the 
electricity has weight and is often compared to inertia for 
purposes of illustration. In the case of the armature in 
question, the sudden shifting from one path to the other 
will induce high E. M. F. 's in the coils, because of the great 
rate of change of current, which tends to maintain the cur- 
rent in the coil just being cut out of circuit and to prevent 
the increase of current in the coil just brought into circuit, 
both of which would cause the current to spark or leap 
through the air from the brush to the segment that has just 



Digitized by VjOOQ IC 



§13 



DYNAMOS AND DYNAMO DESIGN 



17 



left it. This spark, or arc, although perhaps small, is very 
hot and vapo: Izes, or melts, the brushes and segments and 
should be prevented if possible. 

25, A less rapid rate of change of current can be obtained 
by making the commutator segments overlap, after the 



FlO. 9 

manner of Fig. 9, or by using very thick brushes or their 
equivalent, two brushes in 
parallel, as shown in Fig. 10. 
This connects the coils in par- 
allel for an instant, as the 
armature rotates, and the 
brushes are so placed that dur- 
ing this interval the coil com- 
ing into action has a greater 
E. M. F. than that of the one 
about to be disconnected. A 
rush of current through the 
weaker coil is prevented by its E. M. F. of self-induction, 




Pig. 10 



Digitized by VjOOQIC 



18 DYNAMOS AND DYNAMO DESIGN §13 

which opposes the higher E. M. P.. The duration of the 
parallel connection must exist long enough to permit the 
current to die out, but not long enough to allow it to build 
up in the reverse direction, or sparking would again ensue. 

26. Brush Constant-Current Dynanio. — Fig. 9 is a 
diagram of a two-pole Brush armature with two commuta- 
tors connected in series, each of 
which is connected to windings 
exactly like Fig. 10 of Part 1. In 
this figure, the pole pieces are 
represented by the dotted lines 
as they face the sides of the arma- 
I ture. The segments of the two 
separate commutators are, for 
convenience, represented as con- 
centric, with the brushes resting 
on their edges ; whereas, actually, 
they lie side by side, forming two 
separate commutators of the 
^'®- ** same diameter, each having four 

segments, and the brushes rest on their circumferences, as 
shown in Fig. 11. 

One winding consists of two pairs of coils A A' and B B' 
located at right angles to each other, the coils of each pair 
being connected in series, as represented. This winding 
is connected to its commutator, coil A to segment a, coil A* 
to segment a\ coil B to segment ^, and coil B' to segment b\ 
as represented. Brushes 1 and 2 rest on this commutator, 
making contact on the line xy of maximum action of the 
coils. It will be seen that this line is not from center to 
center of the pole pieces, but is moved ahead (in the direc- 
tion of rotation, as indicated by the arrows) from this posi- 
tion by the armature reaction. 

The second winding consists of two pairs of coils C C and 
D D\ located at right angles to each other and half way 
between the coils of the first winding. These coils are con- 
nected in series and to the segments of the second commu- 
tator, coil C to segment c, coil C to segment c\ coil D to 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 19 

segment d^ and coil U to segment d\ as represented. 
Brushes S and ^ rest on the segments of this commutator 
on the same line of maximum action of the coils. 

Taking each winding separately, it will be seen that its two 
sets of coils pass through the following combinations: One set 
of coils only connected to the brushes; then the two sets, 
connected in parallel and both connected to the brushes; 
then one set only ; then both sets in parallel ; and so on. 

The maximum E. M. F. occurs when the single set of 
coils is connected and is directly in the line of maximum 
action; the minimum occurs one-eighth of a revolution 
ahead of this point, when both sets of coils are in parallel 
and are equally distant from the line of maximum action. 

This being the case, it is evident that as the coils of one 
winding are half way between the coils of the other, the 
maximum E. M. F. of one winding occurs at the same 
instant as does the minimum E. M. F. of the other. On 
account of this, when the two windings are connected in 
series, the fluctuations of the current are much reduced. 
This connection of the two windings is obtained by con- 
necting, as shown in Fig. 9, the positive brush of one wind- 
ing with the negative of the other, the external circuit being 
connected between the two remaining brushes. 

In the large sizes of these machines, three, and even four, 
separate windings are used, each with its commutator, and 
all connected in series. In the larger multipolar machines, 
each winding consists of two sets of coils, each set contain- 
ing four coils, one for each pole piece. The action is pre- 
cisely the same as in the bipolar machine. 

27. Tlioinson-Honston Constant-Cnrrent Dynamo. 

Referring again to Fig. 10 of Part 1, it has been found that 
the action of the machine is the same if the two coils are con- 
nected together at the point where they cross each other, 
shown by the heavy black dot in Fig. 12, as these points 
in the two coils always remain at the same potential; the 
resulting winding is shown diagrammatically in Fig: 12, 
which may be considered as having four coils instead of two. 



Digitized by VjOOQ IC 



20 



DYNAMOS AND DYNAMO DESIGN 



§18 



If one of these coils is omitted, the diagram will be that of 
the Thomson- Houston arc dynamo, in which there are but 

three coils — one end of 
each being connected 
together, the other 
ends to each of three 
segments. These seg- 
m e n t s are insulated 
from one another by an 
air space of about 10° 
opening, so the segments themselves span about 110° each. 
Two sets of brushes are used, as in Fig. 10, whose positions 
are automatically controlled by a regulating magnet to 
maintain the current constant regardless of the voltage. 
This regulating magnet is mounted on the frame of the 





PIO. 18 

machine, and the current flowing through it is controlled by 
means of a wall controller, through which the current 
from the machine passes. These machines have been largely 
used for arc lighting and are still found in many of the 
older stations. Under full load, that is, maximum voltage, 
the brushes of like sign are about 60° apart, as shown in 
Fig. 13, while under no load, outside circuit short-circuited. 



Digitized by VjOOQ IC 



§ 13 DYNAMOS AND DYNAMO DESIGN 21 

brushes 1 and S are moved forwards (in the direction of 
rotation) 10° and brushes 2 and Jf, backwards 20**, so that 
those of like sign are about 90** apart. 

In the older machines, the armature is drum-wound, 
although the core is a ring, but in the newer machines a 
ring winding is used ; in either case, three separate coils, or 
sets of coils, make up the winding. 

28. A diagram of the connections, etc. of the drum- 
wound armature is shown in Fig. 13. A A', B B\ and CO 
are the three coils wound on the core one-third the circum- 
ference apart. One end of each of the coils is joined to a 
metal ring (not represented in the figure) on the back of the 
armature, which forms a common connection for the three. 
The other ends are joined to the commutator segments, that 
of ^ -^' to segment tf, that of B B' to segment ^, and that of 
C C to segment r, as represented ; 1 and 2 are the negative, 
and S and U the positive, brushes. Brushes 2 and Jf. are 
usually called the primary brushes and 1 and S the secondary 
brushes^ to distinguish them. 

39. From the diagram, Fig. 13, it will be seen that coil 
A A\ though half way between the pole pieces, is partly 
active, since the neutral line is shifted forwards, in a 
manner that will be taken up later, into the position indi- 
cated by the line xy. This coil A A' is connected in paral- 
lel with the coil B B' by the two positive brushes, and the 
two are in series with coil C C\ If the armature be consid- 
ered as moving in the direction indicated by the arrow, it 
will be seen that as coil A A' gets to the position of least 
action, it is disconnected from the circuit by segment a pass- 
ing out from under brush 3, leaving coil B B' and coil CC 
in series. However, as the distance between brush S and 
brush 2 is only slightly greater than the span of one seg- 
ment, coil A A' is almost immediately connected in parallel 
with coil CC\ as segment a passes under brush 2, making 
the following combination: Coil B B' in series with coils 
A A* and CC in parallel 



Digitized by VjOOQ IC 



22 DYNAMOS AND DYNAMO DESIGN § 13 

As the rotation of the armature continues, coil CO is 
disconnected from the negative brush 1 and connected to 
the positive brush ^ being thus thrown in parallel with coil 
B B\ the two being then in series with coil A A' , Com- 
pleting the half revolution, coil B B^ is disconnected from 
the positive brush S and is joined in parallel with coil A A 
by the two negative brushes 1 and 2, leaving coil C C con- 
nected to the positive brushes. Further rotation of the 
armature repeats this series of connections; that is, during 
every half revolution, one of the coils {A A in the 
preceding paragraphs) is first in parallel with the coil 
behind it, then momentarily disconnected from the circuit, 
then connected in parallel with the coil ahead of it, then 
connected in series with the other two, which are then 
in parallel. 

30, From Fig. 13, it will be seen that when a coil is dis- 
connected from one set of brushes it is very nearly in the 
position of least action, and the coil with which it was just 
before connected in parallel has the higher E. M. F. of the 
two. As explained, the self-induction of the coil prevents 
the higher E. M. F. of the other sending a current through 
it in opposition to its own E. M. F. at the time when they 
are connected in parallel; in fact, when a coil is discon- 
nected from its mate it is still supplying some of the current, 
so that there is a spark at the brushes. 

There being but three sets of coils in this machine, a 
great number of turns must be used in each coil to give the 
required E. M. F., which gives each set of coils a high 
inductance. This lessens to a great extent the fluctuations 
in the E. M. F. acting on the external circuit, which would 
otherwise be very considerable, owing to the small number 
of coils used and the changes in the manner in which they 
are interconnected. 

In the full-load position, only a small fraction of a revolu- 
tion (about 10°) is required to carry a segment from contacts 
with one set of brushes to contact with the other, but if thei 
arc of contact is increased to 70°, or more, the armature 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 23 

will be short-circuited by brushes of opposite sign touching 
the same segment. This short-circuiting occurs at two 
places, i. e., between brushes 1 and ^ and 2 and 5, and, 
there being three commutator segments, six short-circuits 
occur during a revolution. It will be seen that this short- 
circuiting does not reduce the maximum value of the E. M. F., 
but as it introduces periods in each revolution where the 
E. M. F. between brushes is zero, it does reduce its effective 
value. By varying the arc of contact, and thus varying the 
length of time of the short circuit, the effective E. M. F. of 
the machine may be varied between quite wide limits. As 
the field magnets are in series with the armature, their great 
self-induction prevents the strength of the current falling to 
zero, and its fluctuations are therefore comparatively small. 
At the same time, the self-induction of the armature coils 
prevents any excessive flow of current in them during a short 
circuit. 

31. Thus far, the open-coil winding has only been con- 
sidered with reference to bipolar fields. It is evident, how- 
ever, that introducing multipolar fields wilt only result in a 
duplication of the parts used with a bipolar field for each pair 
of poles of the multipolar field. Thus, for a winding for a 
four- pole field equivalent to that shown in Fig. 10 of Part 1, 
four coils would be required in each set. 

Since each set would have to go through its various com- 
binations of connections during each half revolution, instead 
of each revolution, it is evident that twice as many com- 
mutator segments, of one-half the span, would be required, 
or, rather, each segment would be divided into two. These 
two parts of the segment would be situated directly opposite 
each other, and either four brushes would be used on each 
commutator, of which the opposite brushes would have to be 
connected together, or the opposite commutator segments 
would be permanently connected together, and only two 
brushes would be used. This latter plan is best, as perma- 
nent connections are less difficult to maintain than sliding 
connections. 



Digitized by VjOOQ IC 



24 DYNAMOS AND DYNAMO DESIGN § 13 



UNIPOLAR DYNAMOS 

32. Unipolar Induction. — In all the armatures so far 
considered, the conductors as they revolve pass successively 
under poles of opposite polarity. The E. M. F. set up in a 
given conductor, or group of conductors, is alternating 
because the conductors at one instant move under a north 
pole and a moment later they pass under a south pole. If 
arranged so that the armature conductors will always cut 
across the field in the same direction, the E. M. F. and 
current resulting therefrom will always be in the same 
direction and no commutator will be necessary to secure a 
direct current. Such a machine is commonly known as 
a unipolar dynamo ; the term homopolar has also been 
suggested as more appropriate because it is impossible to 
construct a dynamo with only one pole. Unipolar dynamos 
have never been used in practical work except for a few 
special purposes, where a very large current at low voltage 
is desired. However, on account of the introduction of the 
steam turbine for driving dynamos, it is possible that uni- 
polar machines may be used more in the future and the 
principles involved in their operation will, therefore, be 
taken up briefly. 

33. If a copper disk is mounted, so that it projects 
between the poles of a horseshoe magnet, and then rotated, 
the lines of force passing through the disk from pole to pole 
will be cut, and an E. M. F. will be generated. If one con- 
nection is made to the shaft and another to the periphery 
of the disk, by means of a sliding contact, an E. M. F. will 
be obtained between the two terminals. The disk is equiv- 
alent, electrically, to a large number of radial conductors 
connected in parallel, and the E. M. F. generated is the 
same as that obtained from a single conductor only ; how- 
ever, on account of the low resistance of the disk or 



Digitized by VjOOQIC 



§ 13 DYNAMOS AND DYNAMO DESIGN 25 

armature, the machine can furnish a very large current. 
The lines of force are always cut in the same direction, hence 
the E. M. F. induced is direct and absolutely continuous. 

34. In order to obtain any considerable E. M. P., it is 
necessary to use either a very large disk and magnet or to 
rotate the disk at an exceedingly high speed. Many 
attempts have been made to build unipolar machines in 
which a number of disks, or conductors, would be connected 
in series, so as to produce the E. M. F. desired; but all 
these proved failures because the magnetic lines always 
form closed loops, and it is impossible to connect conductors 
in series without the connections themselves also cutting 
the lines in an opposite manner, thus producing an 
E. M F. tl^at exactly neutralizes that which would other- 
wise be obtained by the series-connection. It is possible, 
however, to obtain a high E. M. F. by connecting two or 
more of the revolving disks, or conductors, in series by 
means of sliding contacts, thus making the connecting 
wires stationary and preventing the generation of opposing 
E. -M. F.'s. Heretofore, the E. M. F. that it has been 
possible to generate per conductor has been comparatively 
small and in order to get pressures such as 110, 220, or 
500 volts, a large number of sliding contacts would be 
required, thus making the machine fully as complicated as 
an ordinary dynamo with a commutator. 

The development of the steam turbine, however, affords a 
simple means of procuring high rotative speeds and the 
number of disks can be reduced, so as to make the machines 
practicable in this respect, but there remains the difficulty 
of getting sliding contacts that will operate well at the 
high peripheral speeds of the disks for turbine-driven 
machines. It is difficult to design ordinary direct-current 
dynamos for direct connection to steam turbines, because 
the high rotative speed makes it hard to secure spark- 
less commutation in machines of large output. Unipolar 
machines are free from this trouble because there is no 
commutator. 



Digitized by VjOOQ IC 



26 DYNAMOS AND DYNAMO DESIGN § 13 

36. Fig. 14 shows the essential parts of two unipolar 
machines of the disk type. In {a), a single copper disk a is 
rotated in a magnetic field set up across, the air gap of the 
annular magnet 6 by means of coil c; the path of the mag- 
netic flux is indicated by the dotted lines. In {6), two 



Pig. 14 

disks and two magnets are used and the magnetizing coils 
are connected so that the field passes through one disk in a 
direction opposite to that through the other. The disks 
being in series through the shaft, the E. M. F. between ter- 
minals ^ and e will be twice as great for machine (d) as for (a)^ 
assuming the speed, dimensions, etc. to be the same in both 
cases. In both machines, current is collected by a number of 
sliding contacts projecting through the field frame and bear- 
ing against the sides of the disks near the periphery, as 
indicated at /. Unipolar machines have been built in which 
the armature is in the form of a drum or cylinder, but for 
high-speed machines connected to turbines, the disk form is 
preferable because it can be made of more homogeneous 
material and more accurately balanced. A unipolar machine 
of the type shown in Fig. 14 would have a very low armature 



Digitized by VjOOQ IC 



§ 13 DYNAMOS AND DYNAMO DESIGN 27 

resistance because of the large cross-section of the disks. The 
size of the machine as a whole is determined nearly alto- 
gether by the voltage to be generated, because the disks must 
have a certain cross-section in order to secure mechanical 
strength, and the cross-section thus determined will usually 
have sufficient current-carrying capacity. For these reasons, 
also, unipolar machines cannot be built advantageously for 
small output. In the machines shown in Fig. 14, the disks 
are of copper, but in large dynamos, where heavier disks 
must necessarily be used, the thickness would become so great 
as to insert a large air gap in the magnetic circuit, and 
steel disks are preferable. Where two disks, as in Fig. 14 {d), 
are connected in series, an E. M. F. of 50 volts could be 
generated with disks about 32 inches in diameter run at a 
speed of 2,500 revolutions per minute, assuming that the 
magnetizing coils set up a density of 95,000 lines per square 
inch in the air gap in which the disk rotates. 



CALCULATION OF E. M. F. AND POWER 

36. Calculatloii of E. M. F. — The fundamental princi- 
ple on which all E. M. F. calculations are based is that a 
conductor moving across a magnetic field at a rate such 
that a hundred million, or 10', lines are cut per second, has 
an E. M. F. of 1 volt induced in it. This can be reduced to 
handy working equations for any type of windings. For 
direct-current closed-coil windings the E. M. F. formula 
may be derived as follows: 

Let / be the number of poles of the field magnet ; *, the 
total number of lines of force entering or leaving the arma- 
ture at each pole piece; S, the number of revolutions per 
minute of the armature; Z, the number of face conductors 
on the armature; and w, the number of paths, or circuits, 
through the winding. Then, / ^ lines of force are cut by -a 
single conductor in a revolution, or, in 1 second, each face 

conductor would cut ^^^ lines, so that the average voltage 



Digitized by VjOOQ IC 



28 DYNAMOS AND DYNAMO DESIGN §13 

developed by the conductor would be -^ — rrr^. The voltage 

obtained at the brushes will be equal to the voltage per 
conductor multiplied by the number of conductors in series 
between the brushes. There are Z conductors on the arma- 
ture arranged into a winding with ;;/ paths, and since the 
voltage of all paths must be alike, the windings must be such 
as to have as many conductors in series on one path as on 
another, or, at least, to have the number average the same on 
all paths. So the average number of conductors in series 

Z 

on each path will be — , and as the voltage of the armature E 

is the same as that of one of the parallel paths, then 

60 X 10* ^ nC 

which may be written 

^ p^ZS 

/;/ X 10" X 60' ^ ^ 

which is the fundamental equation for E. M. F. For open- 
coil windings, the average number of face conductors in 

Z 

series should be substituted for — . 

7n 

. Formula 1 is perfectly general for closed-coil windings 

of any kind with either drum-wound or ring-wound coils in 

a field with any number of poles. 

37. Calculation of Povrer. — The number of watts out- 
put of an armature is numerically the product of the E. M. F. 
in volts and the current in amperes'. This number is usu- 
ally divided by a thousand for convenience, and the output 
expressed in kilowatts, abbreviated K. W. Let / be the 
total current in amperes, either entering or leaving the 
armature; this equals the sum of the currents in, say, all 
positive brushes, and let i be the current in amperes in each 
face conductor. Then, since the current divides into m paths 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 29 

equally, I = m i. Now the total electrical power developed 
in the windings in watts is 

Zi is the product of the amperes in each conductor and 
the total number of face conductors, and may be called the 
total armature ampere conductors. / ^ is the product of the 
number of poles and the total magnetic flux from one pole, 
and is sometimes called the total magnetic flux of the 
machine. So it may be stated that 

.^ _ total magnetic flux X total ampere conductors X R. P. M. 
'~ ". 10"* X 60 

(3) 

The quantity / ^ may be expressed in terms of the dimen- 
sions of the armature and of the magnetic density in the air 
gap thus: Let B^, equal the average magnetic density in lines 
per square inch in the air gap between the poles and the 
armature core, and let it, be the percentage of the cylindrical 
surface of the armature core covered by poles ; this percent- 
age is to be expressed as a decimal. Also, let D be the 
diameter of the armature core in inches, and L the length 
of the armature core in inches. Then, nDL is the area of 
the cylindrical surface of the armature core, and ^t: D L is 
the area of the cylindrical surface of the core covered by 
poles; and since the magnetic density in this area is B^, 

The total ampere-conductors Zi in a given line of 
machines usually varies, approximately, with the diameter 
of the armature core, and may be conveniently expressed in 
ampere conductors per inch of diameter or per inch of cir- 
cumference. Using the latter, let K be the ampere conduc- 
tors per inch of circumference of the armature core, then, 

Zi^%DK 



Digitized by VjOOQ IC 



30 DYNAMOS AND DYNAMO DESIGN §13 

Substituting the value of / * in formula 2, gives 

'^' ~ 10' ^ 60 ^ ' 

and substituting in this the value oi Zi = jt Z> AT gives 

^' 10" ^ 60 ^^ 

Thus, the watts output of any armature can be expressed 
in terms of the six quantities ^, Z>, Z,, Bj,, A', and 5, along 
with the constants given. If five of these quantities are 
known, the sixth can be computed. 

38. Considering formula 6, it will be noticed that the 
watts are independent of the number of poles if the other 
quantities are constant; also the voltage, the number of 
paths, and the number of face conductors do not enter the 
equation. If the speed in revolutions per minute is fixed, 
the watts developed will vary with the square of the diam- 
eter of the armature; but if the peripheral velocity instead 
of the speed were limited, this would not be true. 

39. Total Power and Available Output. — The watts 
expressed in formula 5 must not be taken as the output of the 
machine or even the output of the armature itself; they are 
the watts developed within the windings. A small portion 
of the E. M. F. is used in forcing the current through the 
armature resistance and commutator resistance, and if the 
field magnet has a series-winding, there will be a drop in volts 
in that also. If the field magnet has a shunt winding, the 
outside circuit will not receive all the current developed 
in the armature windings, as a part will be us^d in the 
shunt fiekl coils. These losses are known as the electrical 
losses, since they occur only after the electrical energy has 
been developed; the electrical efficiency is the ratio of 
the electrical watts output to the electrical watts developed. 

Let U^ = electrical efficiency of a dynamo, 

W = watts supplied to the external circuit. 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 81 

IV W 

Then. u,= ^otW, = ^^ (6) 

If we call Wf the total power in watts supplied to the 
dynamo by the belt or engine, then the ratio of the watts 
developed JF, to the total watts W^ is called the efficiency 
of conversion f/^. Wf is less than H^, by the watts lost in 
bearing and commutator brush friction, by windage or air 
friction, by hysteresis and eddy currents in the armature 
core, and also by eddy currents in any solid masses of metal 
in which they may be developed. The total efficiency of the 
machine U is called the commercial efficiency. 

^' = f (7) 

or the commercial efficiency equals the product of the elec- 
trical efficiency and the efficiency of conversion. 

Example 1. — 10 horsepower is supplied to the pulley of a dynamo 
and a total E. M. F. of 120 volts is generated in the armature. The 
armature delivers a current of 55 amperes. What is the eflSciency of 
conversion of the dynamo ? 

Solution. — The efficiency of conversion (/c is the ratio of the total 
watts developed in the armature to the total watts supplied to the 

machine; i. e.,C/c= -.jy, formula 7. In this case, the number of watts 

supplied is, W^r = 10 X 746 = 7,460. The number of watts developed 
is 120 X 55 = 6,600; hence the efficiency of conversion 

Ue^ -— = ^'^ = .8847, or B8.47 per cent. Ans. 

Example 2. — In the above example, the loss in the field coils is 
250 watts, and 4 volts is lost in the armature when a current of 
65 amperes is delivered. What is the electrical efficiency of the 
dynamo ? 

Solution. — If 4 volts is lost in the armature, the number of watts 
lost must be 55 X 4 = 220. Hence, the total watts electrical loss is 
220 + 250 = 470, and the watts IV delivered at the terminals of the 
machine will be 6,600 - 470 = 6,130. The electrical efficiency is 

£/; = -^ = ^^ = .9288, or 92.88 per cent. Ans. 



Digitized by VjOOQIC 



32 DYNAMOS AND DYNAMO DESIGN §18 

Example 8. — What is the commercial efficiency of the machine ? 
Solution. — The commercial efficiency is 

i/ = -^ = ?4Ir = -^217, or 82.17 per cent. Ans. 

iVt 7,4oU 

This may also be found as follows. The commercial efficiency 
U^ UcXUe=^ .8847 X .9288 = .8217, or 82.17 per cent. Ans. 

40. Conversion of Mechanical Into Electrical 
Energ^y. — That there is no loss in the actual conversion of 
mechanical into electrical energy may be shown as follows: 
Imagine the armature D inches in diameter and L inches 
long to have ^ of its surface covered by poles with a mag- 
netic density of B^, lines per square inch, and let there be 
i amperes flowing in each of Z conductors, equally spaced 
around the armature. There are ^ Z conductors in the 
magnetic field of a strength of Bj, lines per square inch with 
/ amperes flowing in each. 

The unit current strength has been defined as that cur- 
rent which, when parsing through a circuit of 1 centimeter 
in length bent in an arc of 1 centimeter radius, will exert a 
force of 1 dyne on a unit magnet pole at the center, and this 
unit current is ten times as large as the ampere. The wire 
1 centimeter away from a unit magnet pole is in a field of 
unit strength having a line per square centimeter, or 
6.45 lines per square inch, and the unit current might have 
been defined as that current which, when flowing through a 
straight conductor 1 centimeter long at right angles to a 
magnetic field of one line per square centimeter density, 
caused the conductor to be pushed to one side with a force of 
1 dyne. Inasmuch as doubling the current doubles the 
force, or doubling the field strength doubles the force, it 
may be stated generally that a conductor carrying current 
across a magnetic field has a force acting on it that, meas- 
ured in dynes, is equal to the product of the current strength 
in C. G. S. units times the length of the wire in centimeters, 
times the strength of the magnetic field in lines per square 
centimeter. 



Digitized by VjOOQ IC 



818 DYNAMOS AND DYNAMO DESIGN 88 

Applying this rule to our armature, the current strength 

in each wire is — units, since i is in amperes. The length 

of wire in the magnetic field is L inches for each wire and 

there are i, Z wires in the field, so the length in inches 

is fiZL and in centimeters jiZZx2.54. The strength 

o 
of the field in lines per square centimeter is ^-fz- Hence, 

the force acting in dynes on the rim of the armature 

/ B 

is— X^ZLx 2.54 X ^ dynes. 

Now, these wires are rotated S revolutions per minute, so. 

5 
the inches they move per second would be tt Z> X ^77; or, in 

60 
5 
centimeters, rZ?— - x 2.54. Multiplying the dynes of force 

by the centimeters through which the force acts gives the 
ergs of work; hence, the work in ergs required to drag 
the conductors through the above field for 1 second is 

(± X 5<ZZ X 2.54 X^)(«Z>|x 2.54) 

But 10* ergs make a joule -and a joule per second is a 
watt ; so the number of watts required to drag the above 
conductors is 

{ro X ^^^ X ^-^^ X ^) (.i?| X 2.54)^ 

which reduces to 

watts - ^^-P-^B.XZ.'X5 

"^^"^ " 10- X 60 

This is exactly formula 4, which is the number of watts 
developed in the armature, and the above is the watts 
required to drag the conductors through the magnetic field 
when they are carrying current. Since the two are equal, it 
shows that there is no loss in converting mechanical into 
electrical energy. The losses of a dynamo occur before the 
conversion takes place, as in friction, core losses, windage, 
etc., or after the electrical energy is developed, as in resist- 
ance losses or current to excite the fields. 



Digitized by VjOOQ IC 



U DYNAMOS AND DYNAMO DESIGN §18 



UMITTNG OUTPUT OF CONSTANT-POTENTIAIi 
DYNAMOS 

41. Constant-Potential Dynamos. — The total E. M. F. 
generated by a machine depends on the number of conduc- 
tors, the speed, and the magnetic flux. As all these quali- 
ties are usually kept nearly at the same value at all times 
in closed-coil generators, the E. M. F. of such machines 
remains nearly constant. Dynamos of which this is true 
are called constant-potential maebines, in distinction 
.from constant-current machines, which have already been 
defined. 

The electric power that a machine is developing, usually 
termed its load or output, depends on the product of its 
E. M. F. and the current. To vary the load of a constant- 
current machine, the outside resistance is varied, and with it 
the E. M. F. generated is altered, usually automatically by 

£ 
a regulator, so that -^ will remain constant. With a con- 
stant-potential machine, the current must be varied in order 
to vary the load. This is done by altering the resistance of 

£ 
the external circuit, because from Ohm's law, / = -=; hence, 

if £ is constant, /will vary inversely as R varies. The load 
on a dynamo changes, therefore, with a change in the resist- 
ance of the outside circuit — increasing directly with R in 
the constant-current dynamo until the maximum E. M. F. 
that the machine is able to maintain with the particular 
current the machine is made to give, is reached; while with 
the constant-potential machines, the load increases as the 
outside resistance decreases until the maximum current that 
the machine will safely generate is reached. 

In the constant-current type, the maximum E. M. F. of 
the machine is, of course, fixed by the number of face con- 
ductors on the armature, the speed, and the magnetic flux. 
The magnetic flux is limited by the size and permeability of 
the field magnet, and the amount of wire on the armature is 
limited by the room thereon. A constant-current machine 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 85 

when overloaded is simply unable to maintain the current 
at the particular value designed, and no injury can result 
from such overload. 

43. Factors liimitingr Output. — When the outside 
resistance of a constant-potential machine is greatly reduced, 
an excessive flow of current occurs and the dynamo may be 
injured in two ways: The large current may generate such 
great heat in the armature conductors as to injure them or 
their insulation ; or, the large current may cause sparking 
at the commutator, and thus roughen or burn its surface, 
impairing the further use of the machine. These two limits 
to dynamos are often called the heating: and the sparkinsr 
limits. 

HEATING OF ABMATURK 

43. The heat developed in a conductor is, in watts, 
equal to the square of the current in amperes multiplied by 
the resistance; so, in order that the heat may be limited, 
the resistance of armature windings must be kept very low. 
It is not so much the heat developed in the windings that 
causes the trouble as the resulting temperature, and the 
temperature of a body depends not only on the rate at which 
heat is supplied to it, but also on the ability of its sur- 
faces to radiate, or otherwise get rid of, the heat. Well- 
designed armatures have air passages through the laminated 
cores and also through the windings on the ends to assist in 
dissipating the heat developed, and such armatures are said to 
be ventilated. It will be noticed that the armature heating 
is not only due to the PR loss, as the heat developed in the 
windings is termed, but also to eddy currents and hysteresis 
in the armature core; these last are practically constant, 
regardless of the load, while the PR loss obviously changes 
with the square of the load in constant-potential generators. 

44, It is customary to design modern dynamos so that 
the heating on a 10-hour continuous run with a constant load 
is such that the temperature of the armature is raised about 
40** to 50° C. above the temperature of the surrounding air. 



Digitized by VjOOQ IC 



36 DYNAMOS AND DYNAMO DESIGN §18 

It has been found that insulating material under a constant 
temperature above 60° to 70° C. is slowly destroyed, and 
taking ordinary room temperature as 20° C, this leaves a 
safe rise of from 40° to 50° C. 

Calling the load at which the temperature rise reaches the 
limit set 2iS full load, the load at which sparking will occur 
should not be less than 20 or 40 per cent, over this, and 
preferably at 50 per cent, overload. Many well-designed 
machines will stand about double full load without serious 
sparking. A dynamo thus designed and rated can safely 
stand a continuous load at its full-load value, or a considerable 
overload of short duration or a momentary overload of 50 per 
cent, or more, as the case may be, without any damage 
whatever. 



SPARKING AND COMMUTATION 

46. Commutation is the process that occurs when a coil 
comes into connection with a brush. The primary action is 
the conducting of the current to or from the windings by the 



/ 



Pig. 15 



brush, but in closed-coil armatures this has been shown to 
be accompanied by the reversal of the current in the coil 
during the time of this collection of the current. Under 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 37 

certain conditions this reversal of the current in the coil may 
not be completely or properly accomplished in the time the 
segments pass a brush and undesirable sparking occurs, which 
tends to roughen and otherwise injure the commutator sur- 
face, thus aggravating the difficulty of collecting the current. 
Fig. 15 is a diagram of a portion of a single, parallel, ring- 
wound armature in which the directions of the currents in the 
coils are marked. Observing the direction of these arrows, 
it is clear that as the coils pass from left to right of the brush, 
the current in them is reversed ; it will also be noticed that the 
coils while under commutation are short-circuited by the 
brush. 

46, On account of the self-induction of a coil of wire, the 
reversal, or in fact any change of the current flowing 
through the coil, causes a self-induced E. M. F. that is of 
such direction as to oppose the change of current; further, 
the value of this E. M. F. is proportional to the rate of 
change of current. It is clear that the current cannot be 
instantaneously reversed, otherwise an extremely high self- 
induced E. M. F. would result. The current in a coil when 
short-circuited by a brush does not immediately die down 
to zero, but is maintained by the self-induced E. M. F., unless 
some provision is made to stop it by overcoming this 
E. M. F. 

47. Suppose the brush. Fig. 15, were placed on the 
line O Y\ that is, in a position where the coil under commuta- 
tion is entirely out of the magnetic field ; and suppose the cur- 
rent to be 100 amperes in the brush, or 50 amperes in each of 
the two paths in the winding leading to it. Suppose, further, 
that the coil F under commutation, as shown in Fig. 16, is 
about to pass out from the brush, and that on account of 
the self-induced E. M. F. the current in the coil has only 
decreased from 50 to 40 amperes. Putting the values of the 
currents and their directions on the diagram, it is seen that 
about 90 amperes is forced to flow from segment t to the 
brush through a small and rapidly diminishing area of con- 
tact. As this area is small, its resistance is comparatively 



Digitized by VjOOQIC 



38 



DYNAMOS AND DYNAMO DESIGN 



§13 



high; and as the current density is abnormally high, the 
heat developed at the tip of the brush may be so great as to 
melt the commutator segment and the brush, also, if of 
metal. But this is not all. The segment / an instant later 
passes out from under the brush, and the circuit from / to 
the brush, with 90 amperes flowing in it, is broken. This 
breaking of the circuit will be accompanied by a bright 
spark that is somewhat destructive; and inasmuch as it is 
repeated with each segment at each brush, the trailing tips 
of the brushes are soon burned away, and also the trailing 
edges of the segments. When the circuit from / to the 





Pig. 16 

brush IS opened, a current of 50 amperes must flow from 
right to left through the coil /% and the change in current 
from that shown in the figure is from —40 to +50, or 
90 amperes. If this change is accomplished in a very short 
interval of time, the rate of change will be great and the 
E. M. F. of self-induction will be considerable. This E. M. F. 
will display itself at the terminals of the coil /% or between 
the segments r and /; but since r is in intimate contact with 
the brush, there will be but little potential difference between 
them, and practically the whole of the E. M. F. self-induced 
in the coil F will exist between the brush and the segment /. 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 39 

As the current flowing between / and the brush is throttled 
by the rapidly decreasing area of contact, the E. M. F. of 
self-induction rises; and if it becomes sufficient to maintain 
the current through the air, visible sparking will result. 

It must be remembered that the E. M. F. that causes the 
sparking is that which is self-induced in the coil by the cur- 
rent changing therein, and is not due to the cutting of lines 
of force by the face conductors. 

48. Use of High-Resistance Brushes. — It will be 
observed that since a considerable current is flowing from / to 
the brush through a small contact area having an appreci- 
able resistance, an appreciable E. M. F. will be required to 
maintain this current between the brush and A There can 
be but little E. M. F. between the segment r and the brush, 
as the contact between them is good ; so it follows that the 
E. M. F. between the brush and / is also displayed between / 
and r. This E. M. F. is opposed to the current flowing 
through the brush-contact resistance, because it represents 
the E. M. F. necessary to overcome the resistance, and the 
tendency, therefore, is to divert the current flowing from t 
to the brush and make it take the path through coil F 
instead ; in other words, to reverse the current in F, This 
resistance of the brush-contact area tends to assist in the 
reversal of the current in the coils under commutation, and 
consequently tends to decrease sparking. To make this 
effect more marked, brushes of high specific resistance are 
used — such as graphite, carbon, woven brass or copper 
gauze, all of which are superior in action to leaf-copper or 
strip-copper brushes. 

49. Commutatlng Fringe. — Another way of overcom- 
ing the self-induced E. M. F. is by opposing it by an E. M. F. 
induced in the coil under commutation by its face conduc- 
tors cutting across a weak magnetic field, or fringe, at the 
edge of the poles. The position of the coil under commuta- 
tion is determined by the position of the brushes, and these 
are usually arranged, so as to be adjustable, by mounting 
them on a rocker-arm. 



Digitized by VjOOQ IC 



40 DYNAMOS AND DYNAMO DESIGN §13 

Suppose that the brush, Fig. 15, were rocked toward the 
left until the coil under the brush approached the south 
pole. The E. M. F. induced in it during the short circuit 
would be exactly similar to that in the coils to its left, so 
that the reversal of the current would not be assisted; but 
if rocked forwards to a position such as O X, that is, in the 
direction of rotation, the coil under commutation ap- 
proaches a pole of opposite polarity from that it has just 
left, and the E. M. F. induced by the magnetic field will 
oppose that self-induced during the reversal 

The magnetic field between poles is not zero, except along 
the neutral plane O F, but gradually shades off from posi- 
tive under a north pole to zero and to negative under a 
south pole. It follows, therefore, that for purposes of com- 
mutation, a field of just the right strength may be found 
between the poles by properly adjusting the position of the 
brushes. 

60. It has been stated that the value of the self-induced 
E. M. F. depends on the rate of change of current. Since 
the speed of the machine is usually constant, the time of 
reversal remains constant ; so the rate of change of current 
depends on the value of the current output of the machine. 
It therefore follows that the position of the brushes, as O X^ 
Fig. 15, having been found such that at a certain load 
the reversal of the current is properly accomplished, as 
shown by the absence of any sparking, the position is correct 
for this particular load only, and if the load is greater, the 
brushes must be rocked further forwards and vice versa. 

Dynamos using metal brushes (and some with carbon 
brushes) usually require a shifting of the brushes with load 
variations; and this being undesirable, carbon or other high- 
resistance brushes are now chiefly used on account of their 
additional assistance in reversing the current. In well- 
designed machines, the carbon brushes are set so that the 
current can be sparklessly commutated under full load, and 
it is usually the case that under no load the sparking is 
scarcely visible and not injurious. That the machine -sparks 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 41 

at all under no load is due to the fact that the coil under 

commutation is of very low resistance and is short-circuited 

by a brush while it is in a field and generating an E. M. F. 

Small as is this E. M. F., the resistance of the circuit being 

E 
very small, -^ may be quite large and a considerable local 

current thus induced in the short-circuited coil, only to be 
broken when one of the segments in which it terminates 
leaves the brush. This breaking of the circuit causes spark- 
ing in exactly the same manner as before. 

61. The term sparking is not confined to its literal 
meaning, any improper commutation being given that name. 
Sometimes the current density in the trailing tip becomes so 
high that the carbon brushes may become red hot and glow 
in spots, or the copper in contact with the brush may be 
melted and rubbed off by the brush without visible sparking. 
Sparking may also manifest itself in blackening the com- 
mutator, indicating that the current density is high enough 
to disintegrate the brush or roughen the commutator surface. 

52. Reqalrements for Sparkless Commatatlon. — In 

designing a dynamo, it is necessary, in order to insure good 
commutating qualities, to keep down this self-induced 
E. M. F. This can generally be done by the following means : 

1. Make up the windihg with a large number of armature 
coils, so that the inductance of each may be small. The 
current, will then be reversed in detail, as it were, in the 
windings. 

2. Use a winding having a sufficient number of paths, so 
that the current per path will not be too great. 

3. Use as thick brushes as is practical with a moderate 
speed of commutator surface, so that the time of reversal 
will not be too short. Brushes that are too thick cause 
sparking, because the neutral region is usually quite narrow, 
and if a brush short-circuits a coil while in anything but a 
very weak field, a strong current will be set up in the coil. 
Again, where a number of coils are simultaneously under 



44—9 



Digitized by VjOOQ IC 



42 



DYNAMOS AND DYNAMO DESIGN 



§13 



commutation, they induce E. M. F.*s in each other by 
mutual induction, which also may cause sparking. 

63. In the foregoing, such causes of sparking as rough- 
ness of the commutator, vibration of the machine causing 
the brushes to be jarred from contact with the commutator, 
etc., have not been considered, because these are mechanical 
imperfections in the individual machines, and should be cor- 
rected. It is not always an easy matter, however, to dis- 
tinguish whether sparking in an individual case is due to 
mechanical imperfections or to electrical ones. Often some 
electrical difficulty will cause slight sparking, which may in 
turn so roughen the commutator as to make the trouble 
appear as though the cause were entirely mechanical. 




ARMATUBE REACTION. 

64. By armature reaction is meant the magnetic inter- 
ference of the armature with the magnetic field in which it 

rotates. To thor- 
oughly understand 
the magnetic effects 
of the armature, it is 
best to first consider 
those of a single wire 
in a magnetic field. 
P^o. 17 Suppose that the 

arrows in Fig. 17 represent the magnetic lines of force flow- 
ing between the pole faces of the magnet NS, and let a 
represent the cross- 
section of a wire lying 
at right angles to the 
lines. So long as no 
current flows in the 
wire the magnetic 
field will remain as 

shown; but if the 

' . Fig. 18 

wire carries current, 

say downwards through the paper, the current will tend to 




y 



2 _ V ' i-S!^' ^_ z 




Digitized by VjOOQIC 



§13 



DYNAMOS AND DYNAMO DESIGN 



43 






Pio. 19 



set up lines of force around it, as shown by the dotted 
circles in Fig. 18. It 
will be noticed that 
these lines tend to 
oppose the original 
field below the wire 
and tend to assist it 
above the wire; the 
resultant effect is 
that the field is distorted, as shown in Fig. 19. These mag- 
netic lines act like so many fine elastic bands, with which 
conception the effect of the current in Fig. 19 can be easily 
seen, for the lines of force appear crowded upwards, indica- 
ting that the conductor is pushed downwards, as shown by 
the arrow. 

66. The action of the conductors in the air gap of a 
dynamo is very similar to that described for a single wire, 
except that since there are more conductors their action is 
more marked. If there is no current flowing in the face con- 
ductors, the magnetic lines will cross the air gap radially. 



PIO. 90 



In Pig. 20 IS shown a diagram of the poles and armature of a 
bipolar machine, in which the poles are not magnetized by 
the field coils, but the armature conductors are shown carry- 
ing currents, and magnetism is set up by the armature 



Digitized by VjOOQ IC 



44 DYNAMOS AND DYNAMO DESIGN §13 

currents, as indicated by the dotted lines. In this diagram, 
and in those that follow, a conductor marked with a dot 
indicates that the current is flowing in it upwards from the 
paper, while the solid black circles indicate that the current 
therein is downwards into the paper. 

Applying the rule for the direction of the magnetic lines 
surrounding these conductors, it is seen that each pole piece 
is magnetized so that the lower side is a north pole while the 
upper side is a south pole. The chief reluctance of these 
magnetic circuits is in the air gap, the paths in the iron 
being short ; in the practical case it is not far from the truth 
to consider the reluctance of the iron negligible. It will be 
seen, then, that those magnetic lines which flow from the 
tips of the poles surround or link with the most turns and 
therefore have the most ampere turns displayed along their 
path. If the reluctance of the iron be neglected, the reluc- 
tance of all the concentric magnetic paths or circuits will be 
the same, so that the flux set up will be proportional to the 
number of turns included by the circuit. It is thus seen 
that the strength of the field in the air gap is greatest at the 
tips, reducing to zero at the middle of the poles. 

56. Now imagine the field circuit to be established; 
the combined action of both the armature current and the 
field current will be as shown in Fig. 21. This will be seen 
to be quite similar to Fig. 19, for the magnetic lines in the 
air gap are pushed toward the pole tips c and f by the action 
of the armature currents; the conduct9rs will therefore 
have a force acting on them contrary to the direction of 
rotation indicated by the arrow. Since this force is contrary 
to the direction of rotation, it will require power to drive 
the armature; this mechanical power is absorbed only to 
appear as electrical power in the armature windings. If the 
armature were to revolve in the direction of the force, 
instead of against it, some mechanical force would be 
required to act against that of the conductors so as to keep 
the speed uniform; the armature would then be running 
against this force and would supply mechanical power. This 



Digitized by VjOOQ IC 



§13 t)YNAMOS AND DYNAMO DESIGlf 45 

is the action of an electric motor, and the power obtained 
mechanically must be supplied electrically to maintain the 
armature currents against the E. M. F/sthat will be set up 
by the cutting of lines of force by the face conductors. Not 



PIO. 81 

only are the lines of force in the air gap displaced by the action 
of the armature currents, but they cross the air gap slightly 
on the slant; this produces the pull, or drag, on the conduc- 
tors and the torque, or twisting force, of the armature. 

67. Magnetization at Pole Tips.— In Fig. 21, it will be 
noticed that the strength of the magnetic field varies uni- 
formly from a to c or from e to /. To find the strength of 
field at any point in the air gap as, say, at a, it is necessary 
to take the sum of the strength of field that would be set up 
uniformly over the whole pole face by the field currents, and 
the strength of field shown in Fig. 20. This sum will give 
the combined field due to both field and armature currents. 
It will be noticed that the direction of the field at a. Fig. 20, 
is opposed to that due to the field currents, hence the numeri- 
cal difference of these two field strengths is the strength of 
the field at a, Fig. 21. It is very undesirable to have the 
field very weak at the pole tip, and in practice the field set up 
by the armature currents is never as strong as that due to 



Digitized by VjOOQ IC 



46 DYNAMOS AND DYNAMO DESIGN §13 

the field windings. At the tip r, the field due to the arma- 
ture currents is in the same direction as that due to the field 
currents, so this pole tip is strengthened as shown, being 
numerically the sum of the strengths of the two fields. 

It will be noticed that since, in Fig. 21, the field has been 
shifted, that the neutral region is also shifted and the brushes 
on this account, as well as on account of sparking, should be 
shifted in the direction of rotation, or given a lead, as it is 
called. This will cause the diagram of the face conductors 
to appear as in Fig. 23. 

68. Cross Ampere-Turns and Back Ampere-Turns. 

In Figs. 20 and 21, the currents in the armature conductors 

flow so as to form an 
electromagnet of the 
armature, and magnetize 
it a b o u t an axis a b, 
Fig. 22. As it is imma- 
terial how the armature 
is wound, so far as the 
magnetic action is con- 
cerned, at the instant 
shown it may be consid- 
ered as being made into 
a winding as indicated by 
the dotted lines in the 
figure. Consider Fig. 23, 
in which the brushes have been given a lead of an angle r, 
and are on the line xy. Consider the winding as connected 
up into two coils wound at right angles, as shown, one coil 
having ^ ^ as its axis and the other having A^5 as its axis. 
The coil whose axis is ^ ^ tends to set up a field, as in Fig. 20, 
making the upper side of the armature a north pole, and the 
lower side a south pole. Since this is directly across the 
main magnetic path in the armature, it is known as the 
cross-magnetism, and the turns setting it up, which are 
those included within the angle k, are known as the arma- 
ture ci-oss- turns. The ampere-turns setting up the cross- 
flux, the product of the armature current per wire and the 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 47 

cross-turns, is known as the armature cross ampere-tarns. 
The direction of the currents in the coil whose axis is NS 
is such as to cause its 
ampere-turns to oppose 
those of the field coils, 
and for this reason these 
turns are called th6 
armature back turns; 
and the product of these 
turns and the current in 
each armature conductor 
is known as the armature 
back ampere-turns, or 
counter ampere-turns. 
It will be seen from this 
figure that the angle of 
lead of the brushes r is ^'^" ** 

equal to the angle m ; and further, that from the construc- 
tion, the back turns are those that lie within the angles r 
and m, or within the double angle of lead of the brushes. 
The remainder of the armature turns are the cross-turns, 
and are included within the angle k, 

69. Considering Fig. 20, it will be seen that while all the 
armature turns in this case are cross-turns, the brushes not 
being given any lead, yet all these turns are not available for 
setting up the flux in the poles; only those that lie beneath 
the pole faces are effective in setting up a cross-flux. Those 
that lie between the poles must set up the lines through 
such a long path in air that the flux they produce is practi- 
cally negligible. 

The armature cross-turns may be defined practically as 
those that lie beneath the pole faces. The cross ampere- 
turns tend to weaken the strength of the field under one 
half the pole piece, and tend to strengthen it under the 
other; and since the' weakening effect is equal to the 
strengthening, the total flux on which the voltage depends 
is not changed. That the flux remains the same would be 
absolutely true, were the ampere-turns for the air gap only 



Digitized by VjOOQ IC 



48 DYNAMOS AND DYNAMO DESIGN §13 

to be considered, for the permeability of air is constant. 
But since some of the ampere-turns are required for iron, 
and since the permeability of iron decreases as the density 
increases, the magnetizing effect is usually not quite as 
great as the demagnetizing, because the higher density in 
the iron parts near the strong pole tips requires a greater 
proportionate number of ampere-tums. This will be more 
fully explained in connection with the design of a machine. 

60. For multipolar generators, the action is much the 
same. For both bipolar and multipolar generators with 
drum windings, the total armature ampere-turns per mag- 
netic circuit may be defined as the product of the number 
of turns per pair of poles and the current in each turn. 
Since there are two face conductors per turn, this may be 
expressed as follows : 

Armature ampere-turns = — (9) 

Where Z = number of face conductors; 
/ = current in each conductor; 
/ = number of poles 

This represents the value of the ampere-turns on an 
armature, with the line a b. Fig. 22, as an axis. If this 
quantity is multiplied by the percentage of the armature 
covered by poles ^, the numerical value of the cross ampere- 
turns is obtained thus : 

iZi ,^^, 

Cross ampere-tums = ^-— r— (lO) 

The back ampere-turns are not as prominent as the cross 
ampere-turns, since they are fewer, but they may be com- 
puted thus: 

Back ampere-turns = , (11) 

where r is the angle of lead of the brushes in degrees. 
Applications of these formulas will be shown later in con- 
nection with the designs of a multipolar dynamo. 



Digitized by VjOOQIC 



§13 DYNAMOS AND DYNAMO DESIGN 40 

61. It has been shown that the brushes jnnst be shifted 
ahead of the neutral line in order to bring the short-circuited 
armature coil into a field of sufficient strength to set up the 
proper E. M. F. to reverse the current in the coil. It is easily 
seen that as the armature current increases, the strength of 
the field in which commutation occurs is decreased by the 
cross-magnetizing force, so that the brushes must be shifted 
still farther ahead to reach a field of the proper strength. 
This increases the angle of lead and the back ampere-turns 
become greater and greater, further weakening the field; if 
the armature current is great enough, the brushes might be 
shifted any amount without reaching a sufficient field for 
commutation. The armature reaction then limits the 
amount of current that the armature can supply without an 
excessive shifting of the brushes. This limit of the output 
is the sparking limit. 

62, General Eflfects of Armatnre Reaction. — Arma- 
ture reaction does three things : 

1. It inclines the magnetic lines in the air gap, producing 
the torque or twisting force. This is absolutely essential 
for the proper action of the machine. 

2. It shifts the field in the direction of rotation in a 
dynamo, due to the cross-magnetizing force of the ampere- 
turns beneath the pole faces. This effect may be lessened 
by increasing *the reluctance of the path of the cross-flux. 
One of the chief methods of doing this is by saturating 
magnetically the pole tips and armature teeth, as will be 
fully explained in connection with the design of a generator. 
The cross ampere-turns may be neutralized by placing on 
the poles, windings in series with the armature winding and 
arranging them to set up a flux in a reverse direction to 
that of the armature winding. 

3. It tends to weaken the field strength due to the coun- 
ter-magnetizing forces. This effect is easily neutralized by 
placing on the magnet cores series-windings of the proper 
number of turns. 



Digitized by VjOOQIC 



60 DYNAMOS AND DYNAMO DESIGN §13 



COK8TBUCTION OF THE AllMATUBE 



CONSTRUCTION OP CORE AND SPIDER 

63. The armature core, as it is called, consists usually 
of a number of thin punchings of a special grade of iron, or 
steel, selected especially for the purpose on account of its 
having high permeability and also being subject to only a 
moderate loss per cubic inch by hysteresis. Occasionally, 
fine iron wire and iron strips are used, but these are so rare 
as to be hardly worthy of notice. The punchings are usu- 
ally from .014 to .03 inch thick, although .06 inch has occa- 
sionally been used. As has already been explained, this 
lamination of the core is necessary to prevent the forma- 
tion of eddy currents within the core; and since the loss 
caused by eddy currents varies with the square of the thick- 
ness of the punchings, it is evident that thin stampings are 
much better than thick ones. The losses in punchings less 
than .02 inch, or 20 mils thick, are so small that the addi- 
tional cost of using much thinner material does not result 
in a gain sufficient to compensate therefor. 

These punchings must be supported on a shaft. They 
may be slipped directly thereon or they may be mounted on 

a spider, which, in 
turn, has a hub 
through which the 
shaft is pressed. In 
[ bipolar machines, 
the flux passes from 
diametrically oppo- 
site points through 
the core; therefore 
Fia u it is best to have as 

small a central hole in the punchings as possible; they are 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 61 

in this case simply slip- 
ped on the shaft, as 
shown in Fig. 24. The 
punchings are held 
^ firmly in place between 

two end plates, a^ a, usu- 
ally of cast iron, which 

§ are clamped between a 

shoulder c on the shaft 

and the nut 6. To pre- 

^ vent the disks from 

turning, a key inserted 
in the keyway d pro- 
jects above the shaft 
into a suitably cut nick 
in each punching. The 
nut b need not be hex- 
agonal; it is often 
^ round, and can be well 
I screwed up by a pipe 
wrench. In another 

^ construction the nut is 

replaced by a tight 
sleeve, which is pressed 
on with a hydraulic or 
other press. 

64. Fig. 25 shows 
the construction of a 
multipolar armature in 
which the disks are 
punched, as shown at 
^ {^), and slipped on the 
"^ shaft between two end 
plates £, £ {c), one of 
which is held by the 
shoulder /; the other is 
preferably pressed on 



Digitized by VjOOQ IC 



62 DVNAMOS AND DYNAMO DESIGN | IS 

and held by a nut n on the shaft. It will be noticed 
that the path of the flux from pole to pole is along a 
short cord in a multipolar armature, rather than along a 
diameter, as in the bipolar; and, further, the poles in the 
former being smaller with the same diameter, a large radial 
depth of iron x is not required to convey the magnetic flux. 
In other words, using the same diameter of core disks, the 
dimension x may be decreased as the number of poles is 
increased. The holes k are therefore punched out, leaving 
the arms k for supporting the rim, which, with the teeth, 
constitutes the real armature core. This construction greatly 
lightens the armature and permits the circulation of air 
through the core to assist in carrying away the heat devel- 
oped therein. This keeps down the rise in temperature 
or, with a given rise in temperature, permits a greater load. 
The end plates E^ E should have a long hub to firmly attach 
them to the shaft. At {c) these end plates are shown with 
extensions for supporting the windings Wy IV. 

66. Armature Spiders. — The construction shown in 
Fig. 25 saves somewhat in labor by avoiding the use of a 
spider, but it is more wasteful of armature Iton, since the 
central part of the disks, if punched out in a single piece, 
would doubtless be used for smaller armatures It is also 
objectionable in that the shaft cannot be removed without 
disturbing the punchings. 

When dynamos are direct-connected to engines or water- 
wheels, it is customary to supply the armature complete 
without the shaft, the engine builder supplying this with the 
engine; it is desirable, therefore, to be able to -build the 
armature without a shaft. 

66. An armature core intended for a ring type of wind- 
ing is shown in Fig. 26. The disks are supported on a four- 
arm spider made in two parts, which are clamped together 
by the nut n on the shaft. For ring windings, the spider 
arms s, s' are made very thin, say from ^ to ^^ inch thick, 
so as not to interfere with the windings on the inside of the 
core. To each part of the spider, an end ring r, r' is 



Digitized by VjOOQ IC 



§13 



DYNAMOS AND DYNAMO DESIGN 



58 



attached; these clamp the disks together; the arms project- 
ing beyond the rings in the form of wings are merely added 
to obtain additional strength. A small key k in one of the 
spider arms prevents the disks from turning on the spider, 
while another key k' prevents the spider from turning on the 
shaft. This style of spider "has been used for ring armature 
for arc-light dynamos. One objection to it is that the hub 
is in two parts, which interferes to some extent with remov- 
ing or inserting the shaft, after the armature has been built. 




PIO. 96 

67. In order that the disks may be readily slipped over 
the spiders, which is done before they are put together, each 
spider is provided with two lengths of arm. Opposite arms 
are of the same length, as seen in full lines in the cross- 
sectional view of Fig. 26, but two are shorter than the other 
two, as is indicated by the dotted lines at a. Each spider may 
thus have a number of disks laid over it, up to the full extent 
of the longer arms, and the two spiders are then pressed 
together with clamps. It is better to leave a small clearance 
space between the ends of opposite arms, and allow the hubs 
to meet on the shaft. After the disks are all in place, the 
spiders containing them are pressed on the shaft, and the 
nut n is screwed up to hold them. A setscrew / should be 
used to prevent the nut from loosening. 



Digitized by VjOOQ IC 



64 DYNAMOS AND DYNAMO DESIGN § 18 




Digitized by VjOOQIC 



§ 13 DYNAMOS AND DYNAMO DESIGN 65 

In small, ring-wound armatures the spiders are usually 
made of bronze, to prevent leakage of the magnetism 
through them across from pole to pole, for such flux would 
set up in the wires on the inside of the core E. M. F.*s 
directly opposed to those induced in the face conductors. 
On larger' machines this is hardly important enough to 
warrant the expense of large bronze castings. 

68, Fig. 27 shows an armature core mounted on a spider, 
which can be removed from the shaft without taking the 
armature apart. Here the spider consists of two parts, the 
spider proper and a small end ring c for clamping the disks, 
which is attached to the spider by tap bolts d. Instead of 
using a key for keeping the disks from turning, they are 
punched with a tongue shown at a, which fits into a keyway 
cut into one of the spider arms. The arms are wider and 
thicker than in Fig. 26, for the winding being of the drum 
type does not pass through the center of the core and is 
therefore not interfered with by heavy arms. The end 
rings on the spider, for clamping the disks, are strengthened 
by the rib b^ which runs completely around, as shown. 

In the section, the winding h is shown in place, the end 
connections being held and protected by the flanges /, /. It 
must not be thought that the section represents the shape 
of a coil, for the two sides of a coil (upper in one slot, lower 
in another) must be separated by the angular distance 
between corresponding points on adjacent poles, as has 
already been explained. 

Aty, the hub of the spider is shouldered down to receive 
the commutator, which is of the same style as that shown 
in Fig. 39. It will be seen that the completed armature is 
independent of the shaft, so that this construction is espe- 
cially suited for direct-connected machines. 

69, An armature core and spider for a large, barrel- 
wound, electric-railway generator are shown in Fig. 28. 
The spider consists of a hub and radial arms only, both the 
end plates being separate. The hub is cored out so as to 
leave only three points for boring out in order to reduce the 



Digitized by VjOOQ IC 




Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 57 

labor of machining. As in Fig. 27, the commutator spider 
is fitted on to an extension of the hub of the armature 
spider. The spider arms are elliptical in section, as shown 
at ^, and are provided with heavy end-pieces having 
pockets c into which lead may be swaged for the purpose of 
balancing the armature. The armature end plates /?, ^are 
provided with flanges/, /that fit under the end-pieces of the 
spider arms. The end plate D has a flange g^ which, being 
as large in diameter as the armature core, protects the wind- 
ings supported by it from injury. 

The armature core disks are made in five segments, the 
diameter being too great to admit the use of a single punch- 
ing per disk, as is usually the case with smaller machines. 
Each segment has two lips, or dovetail projections, on its 
inner edge, which fit properly into recesses in the spider arm, 
as shown, the whole core being clamped firmly together by 
the bolts h. In assembling the disks, the joints between 
segments are broken ; that is, if on one disk the joints are 
at k^ k^ on the next they will be at /, /, etc. 

70, Ventilating Ducts. — It will also be noticed in the 
section of the core that openings v^ z\ i\ Fig. 28, are shown 
between the disks. These are called ventilating: ducts, or 
flues, and are air passages in the core formed by properly 
made spacers. These ducts are for the purpose of permitting 
air to be forced outwards through th'em by the centrifugal 
action, thus radiating the heat due to the armature losses. 
This causes the rise in temperature for a given load to be 
very much lessened, and therefore raises the heating limit of 
the machine; that is, it increases the load at which the 
machine will rise in temperature to such a degree as to 
endanger the durability of the insulating materials. 

71. 8niootli-Core and Toothed Armatures. — Arma- 
ture cores are referred to as smootli-core, and slotted or 
tootlied, according to whether the face conductors are upon 
the outside cylindrical surface of the core or are held in 
slots. Windings placed upon smooth-core armatures are 
usually called surface wlndlugfs. In modern practice, 

44—10 



Digitized by VjOOQ IC 



58 



DYNAMOS AND DYNAMO DESIGN 



§13 



these are not much used because the output of a dynamo 
may be increased by using a slotted armature and also 
because the windings in the latter case are thoroughly pro- 
tected and can be better insulated. 

In winding a smooth-core armature, the core should first 
be carefully insulated and stout fiber driving pegs driven 
into properly cut recesses in the end plates, or core proper, 
for the purpose of guiding the winding and also for driving 
it against the reaction of the magnetic field. 

72, Armature Slots. — Armature slots are of a variety 
of shapes, some of which are shown in Figs. 29 and Figs. 30. 
At (tf). Fig. 29, is shown a plain straight slot with parallel 




M 



Pio. 90 




sides. This is the simplest and perhaps the most used of any. 
At (d) and (c) are shown modifications having nicks at or 
near the top into which fit retaining wedges of hard wood or 
fiber. These wedges not only protect the winding from 
injury, but take the place of wire bands, which are otherwise 
necessary, for retaining the winding in place against the 
centrifugal force due to rotation. Of course, on armatures 
using retaining wedges, band wires must be used on the end 
connections beyond the core. 

yiTlT7 lUT ^m 



(•) 



Pio. 80 



(0 



In Fig. 30 are shown some of the shapes of slots in 
which the teeth oVerhang. These types are usually resorted 



Digitized by VjOOQIC 



§13 



DYNAMOS AND DYNAMO DESIGN 



59 



to in order to keep the opening of the slot small and still 
have a liberal size of slot. A wide opening at the surface is 
objectionable when the length of the air gap between the 
pole face and the teeth is small, since in that case the mag- 
netism passes across the pole face in tufts, see Fig. 31, and 



TTrrnnnmz 





Pig. 81 

is then liable to set up eddy currents in the pole face because 
of the rapid passage of magnetic flux of varying densities 
across the face. When the losses incurred by these currents 
are objectionable, the pole pieces may be laminated, like the 
armature core, in which case the slot openings may be very 
wide without occasioning any material loss. 

73. In this connection it may be stated that it has been 
found that where the average air-gap magnetic density is 
less than 35,000 lines per square inch, solid poles may be 

used if the length of air gap A is greater than — , but if riot, 

o 

the pole pieces should be laminated. If the air-gap density 
is about 50,000 or more, A should be greater than — with 

solid poles. 

Sometimes the slots are entirely closed at the tops and 
become holes near the periphery of the core. This construc- 
tion thoroughly protects the windings, and does away with 
all liability of pole-face eddy currents, but is very objection 
able, because the armatures are difficult to wind in com- 
parison to the open-slot type, and also because the self- 
induction of the coils is so greatly increased that it is diffi- 
cult to prevent commutator sparking. 



Digitized by VjOOQIC 



60 



DYNAMOS AND DYNAMO DESIGN 



§13 



METHOD OF APPI.YING WINDINGS 

74, Construction of Colls. — The usual shapes of arma- 
ture coils have already been shown. The barrel type of 
winding is now generally used, for it spreads the windings 
out on a rapidly moving surface and thus readily dissipates 
the heat developed by the I^R losses. Coils of this shape, 





where wound with wire of several turns, may be wound 
directly upon a frame, or form, having the peculiar shape of 
the finished coil, or they may be wound into a single loop, as 
shown at (tf), Fig. 32, and afterwards put into a frame and 

pulled out into the shape 
shown at {b). Where 
the coil is made of a 
single turn of copper 
strip or of heavy round 
copper, the same proc- 
ess may be used, but 
the pulling-out frames 
are not satisfactory for 
heavier conductors. In 
forming coils of copper 
strip of, say, from | inch 
to IJ inches in width, pieces of the copper may be bent as 




Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 61. 

shown at (^), Fig. 33, then insulated and assembled on the 
armature, after which the proper ends are soldered together 
by a clip at d, forming a coil as shown at (b). In Fig. 33, 
parts aa, cc lie in the slots, while parts bb form the end 
connections. Some makers avoid the soldered joint at d by 
first bending the strip on edge into a long U; it is then 
clamped into a suitable cast-iron former, and the annealed 
copper hammered to the required shape with a mallet. 
When the coil is completely shaped the clamps are removed 
and the coil carefully taken out, so as to preserve its correct 
shape. It is then wrapped with cotton, or sometimes linen, 
tape J inch wide, treated with armature varnish — a prep- 
aration consisting chiefly of linseed oil — and then baked in 
an oven at a moderate temperature. 

75, Slot Insulation. — Before inserting the coils in the 
armature slots, the latter are insulated with a cell of mica, 
presspaper, or pressboard, oiled 
muslin, or a combination of these 
in the form of a thin sheet usu- 
ally from .02 to .06 inch thick, 
shown at b, Fig. 34. The coil is 
then inserted in the slot and over 
it is placed a layer a of insulating 
material usually a little thicker 
than the cell. The upper coil 
is then inserted, the cell bent 
down over the upper coil, see 
Fig. 35, and finally the hard- 
wood or fiber wedge pushed 
into place. It is, of course, 
evident that there are many 
satisfactory ways of insulating 
the coils, but the above serves 
as a typical example. It is 
very similar to the method ^^^' ^ 

used by several prominent manufacturers on their larger 
machines. 



Digitized by VjOOQIC 



•62 DYNAMOS AND DYNAMO DESIGN §13 

76. For low-potential generators, say 125 or 250 volts, a 

very good slot insulation may be 
_ made up from two thicknesses 
of thin presspaper .01 inch 
or .015 inch thick with a layer 
of oiled muslin between them, 
the whole being stuck together 
with shellac ; if this is bent into 
shape before it is dry, it is readily 
handled. Such an insulation is 
called a sandwicli. If the poten- 
tial of the machine is greater, 
very thin leaves of mica may 
be added between the sheets 
of presspaper until the desired 
insulating qualities are obtained. 
On arc-lighting generators, 
troughs, or cells, of mica, or 
micanite — a preparation of thin 
^^' ^ sheets of mica stuck together 

with shellac or other varnish— are used, but for potentials of 
600 volts or under, such slot insulations are unnecessarily 
expensive. 



SHAFTS 

77, Dynamo sliafts must be somewhat stifFer than shafts 
for general machinery. The air gap of a dynamo is usually 
from ^ inch to J or J inch, and if the shaft springs or 
deflects under the armature weight, the armature will become 
eccentric in the bore of the poles and thus shorten the air 
gap on one side and lengthen it on the other. The pole with 
the shortest air gap will naturally set up the greatest flux 
and therefore cause the greatest pull on the armature toward 
it. This exaggerates the trouble, for the excessive magnetic 
pull may cause the armature to be still further deflected 
from its central position and it may ultimately strike the 
pole pieces. 



Digitized by VjOOQ IC 



§13 



DYNAMOS AND DYNAMO DESIGN 



63 



The diameter of a shaft for a given dynamo may be 
computed on either of the following rules with satisfactory 
results: (1) The deflection of the shaft due to the weight of 
the armature and commutator should be between the limits 
.004 inch to .008 inch at the face of the armature. (2) The 
maximum fiber stress in the metal should be between the 
limits 5,000 to 8,000 pounds per square inch. The following 
formula, however, may be used for getting the approximate 
size of a shaft : 



diameter 






watts 



P.M. 



(12) 



In which k may be taken from the following table for 
belted dynamos. For engine-typ)e dynamos shafts are 
special and are usually made by the engine builders, the 
bore of the armature spider being made to suit. 

TABIiE I 
VALUES OF k rN^ FORMULA 1« 



Kilowatts 


k 


Kilowatts 


k 


I 


.90 


50 


1. 10 


5 


•95 


100 


1.20 


ID 


1. 00 


500 


1.40 



78, Having obtained the diameter by this formula, the 
stresses should be checked up, not forgetting those due to 
the pull of the belt, and the deflection of the armature core 
proper should be computed. The shaft should be made at 
least as large as the greatest value required by the limits 
adopted by the rules. Methods of computing the deflection 
or the maximum fiber stresses cannot be taken up at this 
point ; they belong properly to the subjects of mechanics 
and machine design. 

The shafts must be shouldered down to meet the require- 
ments in the design, so that a shoulder or collar is provided 
at or near the inner side of each bearing to limit the endwise 



Digitized by VjOOQ IC 



64 



DYNAMOS AND DYNAMO DESIGN §13 



motion of the shaft in the bearings. This end motion may 
best be limited in this way, for it is only at the bearings that 
wearing surfaces are supplied with oil. 



BEARINGS 

79, The size of bearlngps for dynamos is determined in 
the same manner as are other bearings, except that the con- 
stants for the formulas used are selected so as to give rather 
larger bearings than those for general machinery. Bearings 
for dynamos usually have a length of from three to four 
times their diameter. 

Dynamo bearings are always made of the ring-oiling type; 
that is, the bearing is supplied with a reservoir for oil into 
which dip one or more rings that run upon the shaft, and 
thus supply the journal continually with oil. 



COMMUTATORS 

80, The commutator, as has already been explained, 
consists of a number of copper segments clamped together 
by suitable clamping rings, each segment being insulated 
from the others and from the clamping rings by mica or 
micanite. The brushes that serve to convey the current to 
or from the commutator rub on its outside cylindrical sur- 
face, which is turned true and smooth for this purpose. 

Figs. 30, 37, and 38 illustrate three commonly used shapes 
of segments and their clamping devices. Fig. 36 illustrates 
the form of commu- 
tator used on the 
Edison bipolar dyna- 
mos, an early type 
of exceptional merit. 
It consists of a shell, 
or sleeve, s on which 
is screwed the ring/. 
A micanite cone, or 
nng, / IS slipped on, 
then the mica sleeve m, the copper segments and the mica 




Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN * 65 

insulating segments, the mica cone /, the wedge ring h, and 
finally the nut k is screwed up, clamping the whole firmly 
together. It is nec- 
essary that the com- 
mutator be very 
carefully made i n 
order that it may 
successfully stand 
the strains of 
machining it and 
be perfect electri- 
cally. To this end 
the segments are 

very carefully ^'°- ^ 

tapered from the outer to the inner edge, to within less than 
YsVy inch of the correct size, and the mica segments are cali- 
pered and made to a similar degree of accuracy. Shellac is 
liberally used in putting the commutator together, and after 
it is clamped, it is usually heated and screwed as tight as 
possible, cooled, and again tightened. On account of the 
very severe strains of tightening, the shell is subjected to 
very heavy strains and must therefore be of ample strength. 
If the shell is of iron, it may also be subjected to very severe 
strains on account of the copper bars expanding more than 
the iron shell when heated. Bronze is used for the shells of 
small commutators and cast steel or malleable iron for mod- 
erate-sized ones. If the wedging strains are taken by bolts, 
the shells may be of cast iron, especially in the larger sizes, 
where the tightening strains are not proportionally as large 
as in the smaller commutators. Note the method of con- 
necting the conductors to the segment: The leads r from 
the two coils are first soldered firmly into the terminal /, 
which in turn is attached to the neck of the segment w by 
the two screws. In Fig. 37, the segment has but a very 
short neck, in which is milled a slot just wide enough for the 
conductors r, which are directly soldered in, thus avoiding 
the extra terminals of Fig. 36. This is a better and more 
common method of attaching the leads. 



Digitized by VjOOQIC 



66 DYNAMOS AND DYNAMO DESIGN §18 

81. The style of segment shown in Fig. 37 is quite satis- 
factory and is largely used. The diameter of the face, or 
working surface, is much greater than that of Fig. 36, as 
both the tightening bolts and clamping rings are beneath 
the surface of the face. This larger diameter is, of course, 
much more expensive in material and labor, but it is neces- 
sitated by the use of carbon brushes and multipolar types 
of field, as against metal brushes and bipolar fields. Metal 
brushes have low resistance, and therefore are small, while 
the heat developed in the face of the commutator due to 
both PR loss and to friction of the brushes is small, so that 
a commutator with a small surface for dissipating the heat 
is satisfactory. On the other hand, carbon brushes are of 
high resistance, a property useful in keeping down sparking, 
and must therefore be large, so the friction and PR losses 
are greater and the area for radiation must be increased in 
order that the commutator may not overheat. As it is, the 
commutator is usually the hottest running part of a modern 
dynamo, for the heat there can do little harm aside from pro- 
ducing strains in the structure due to unequal expansion of 
the metals, unless the temperature rises to such a degree as 
to melt the solder used in attaching the leads from the coils. 
The rise of temperature of a commutator is often limited 
to bb^ C. above that of the surrounding atmosphere. 

82. Multipolar designs require a larger diameter of com- 
mutator than bipolar designs, because they require more 
segments in proportion to the number of poles. It has been 
shown that, from one neutral point on the commutator to 
the next, there exists the full difference of potential of the 
generator, and the average number of volts per segment 
cannot be greater than a certain limited quantity. The 
mica between the segments cannot be safely made less than 
from .02 inch to .025 inch thick without causing trouble; 
and further, it has been found that a commutator with more 
than from 12 to 15 per cent, mica in its make-up will not 
wear well, and is therefore undesirable. Thus, if we limit 
ourselves to .025 inch mica as the thinnest insulation and 



Digitized by VjOOQ IC 



§ 13 DYNAMOS AND DYNAMO DESIGN 67 

12 per cent, as the greatest amount of mica, then each seg- 
ment and mica must be not less than *— — - = .2 inch. This 

is not a bad limit to keep within; and for high voltages, 
small-sized multipolar generators will nsually call for a fairly 
large diameter of commutator. 

83. The segment a. Fig. 37, is clamped between the 
shell 5 and the wedge ring //, the two being clamped together 
by the bolts b. This construction, by putting the clamps 
below the surface, at once increases the diameter and 
shortens the length. For voltages of 250 or over, the mica 
cones /, / are preferably made longer than the segments, 
and a layer of string is carefully wound over the projecting 
part to protect it, as shown. This is done in order to keep 
the current from arcing over to the frame and thus ground- 
ing the machine. 

84, Fig. 38 is a type of commutator suitable for large 
generators. The segments have a very long neck to make 
up for the differ- 
ence in the diame- 
ters of the armature 

and commutator, 
the armature con- 
ductors here being 
too large to bend 
down to the seg- 
ment. The commu- 
tatjor consists of a 
spider b with a hub, 
not shown: two 
clamping rings c^ c, 
made in segments, 
and held in place 

by the bolts/; and two wedging rings d, made in segments. 
It will be seen that if the tap bolts e- are tightened, ^ will be 
drawn up, wedging the clamps c inwards and thus clamping 



Digitized by VjOOQIC 



68 



DYNAMOS AND DYNAMO DESIGN 



§13 



the commutator segments a. The rings c must evidently 
be made in segments. It is claimed for such a construction 
that defective copper segments may be replaced by removing 




only the segment of c, under which the defective part is 
held. Such a commutator has a very good property of 



Digitized by VjOOQ IC 



§ 13 DYNAMOS AND DYNAMO DESIGN 69 

permitting air to circulate through its interior, thus assist- 
ing in cooling it; but it will be noticed that the heat is all 
developed in the copper segments that are separated by the 
mica from the spider, and mica is a fair heat insulator as 
well as electric insulator. Another good quality of this 
design is that the clamps ^, c need not be tightly drawn 
together to hold the segments, as in Figs. 36 and 37, so the 
expansion and contraction of the copper segments can be 
permitted. 

85, A completed commutator, of the type used with the 
armature shown in Fig. 27, is shown in Fig. 39. It is very 
similar in style to that shown in Fig. 37, its peculiarities 
being the use of tap bolts instead of through clamping bolts, 
and also the use of setscrews in the segments for binding the 
conductors. The outside appearance is improved by the use 
of tap bolts, but they are generally more expensive than 
through bolts. 

Fig. 40 shows a per- 
spective and a half sec- 
tion of a commutator of 
a railway motor. It is 
a slight modification of 
Fig. 37, in that its diam- 
eter is not great enough 
to permit bolts being 
used beneath the seg- 
ments, so a special nut e 
is used, which is 
screwed upon the shell 
by a spanner fitting into 
the holes shown, a set- 
sere w being used to 
prevent its loosening ; 
/ is the bar, ^, b the 
mica insulating cones, a the shell, and d the clamping 
ring. 



PIO. 40 



Digitized by VjOOQ IC 



70 DYNAMOS AND DYNAMO DESIGN 813 



ABMATURE IX>SSES AND HBATTNG 

86. Since the output of a dynamo is limited by the tem- 
perature rise of its various parts, it is important that the 
heating be carefully considered in designing. The tempera- 
ture rise depends on two factors, the rate at which heat is 
developed within the body and the ability of the body to 
dissipate the heat. The former quantity, usually termed 
the losses, may be computed with reasonable accuracy, but 
the latter is usually best determined by experience. 

87. Division of Xiosses. — The losses in aii armature may 
be divided into first the copper losses, which occur within 
the conductors due to their resistance ; and second the iron 
losses, namely those due to hysteresis and to eddy currents 
in the armature core. Since the numerical value of the 
watts lost in a conductor due to its resistance is the product 
of the square of the current in amperes and the resistance 
of the conductor in ohms, the loss is commonly spoken of as 
the /• R loss. This is frequently referred to as the C* R loss, 
since the letter C was formerly used to denote current. 

88. Estimation of Armature Resistance. — To com- 
pute the PR loss of a given armature, since its current 
output is known, it is merely necessary to determine its 
resistance in ohms. To measure the resistance of the wind- 
ings of an armature of any size is a matter of considerable 
difficulty, because the resistance of the armature itself is 
usually so small that the resistance of the terminals is quite 
a large part of the total, and it is difficult to eliminate this 
from the measurements. Consider for instance a parallel or 
multicircuit winding for a multipolar generator. To obtain 
the resistance of all paths in parallel, since they cannot be 
individually measured in a winding conipletely connected 
up, they must be connected together as they are when in 
operation, and these connections involve the brushes, brush- 
holders, and their connecting wires. However, it is not 
often necessary to measure the resistance of such a winding, 
as reliable results can be obtained from calculations. 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 71 

Obviously the number of circular mils area of a round 
wire is obtained by squaring its diameter in mils. 

It has already been shown that the resistance in ohms of 
any copper wire of uniform cross-section at ordinary tem- 
peratures may be computed from the formula 

;p — 10-8 X length in feet . .^^x 

"" area in circular mils ^ ' 

In dynamo calculations the resistance of a mil foot of wire 
will be considerably above 10.8 ohms, because the wire is 
warm when the machine is in operation. A fair value for 
the resistance of a mil foot under these conditions is 12 ohms; 

^' • * o 12 X length in feet ,^-. 

warm resistance R = : ? — -. tt- (14:) 

area in circular mils "^ ' 

and since there are 12 inches in a foot, this may be written 

r> __ length in inches ,^ -. 

~" area in circular mils ^ ' 

89, To compute the resistance of an armature coil, it is 
necessary, first, to estimate its length and the area of the 
section of copper. If the copper conductors are of rectan- 
gular section, the area in square mils may be obtained by 
multiplying together the two dimensions of the wire in 
thousandths of an inch. A circular mil has a smaller area 
than a square mil by the ratio of the area of a circle to that 
of a square, or by .7854; so that if the area of a conductor in 
square mils is divided by .7854, the result will be the area in 
circular mils. 

Let r^ be the resistance of a single armature coil calcu- 
lated, say, from formula 15, and let C be the number of 
coils on the armature connected up into a winding having 
m paths, or circuits. Then the resistance of a single path, 

C r C 

since there will be — coils in series per path, is — — . But 

there are m such paths in parallel in the complete armature 

winding, so the total resistance will be only — of that of one 

path, or r ^ C 

K = ^^ (16) 



Digitized by VjOOQIC 



72 



DYNAMOS AND DYNAMO DESIGN 



§13 



TABIiE n 

MAGNET WIRE, BROWN A SHABPE GAUGE 





Diameter in Mils 


Area 

Circular 

Mils 


Ohms Per 


Pounds 
Per 
1,000 
Feet 


Gauge 
Number 


Bare 


Single 
Cotton- 


Double 
Cotton - 


Triple 
Cotton- 


1,000 
Feet at 

20»C. 






Covered 


Covered 


Covered 




68«F. 


Bare 


oooo 


460.0 






478 


211,600 


.049 


640.50 


ooo 


409.6 






428 


167,805 


.062 


508.00 


oo 


364.8 






383 


133.079 


.078 


402.80 


o 


324.9 






343 


105.534 


.098 


319- 50 


I 


289.3 






307 


83.694 


.124 


253.30 


2 


257.6 






276 


66,373 


.156 


200.90 


3 


229.4 






247 


52,634 


.197 


159.30 


4 


204.3 




216 


220 


41.742 


.248 


126.40 


5 


181. 9 




194 


198 


33.102 


.313 


100.20 


6 


162.0 




174 


178 


26,250 


.394 


79.46 


7 


1443 




156 


160 


20,816 


.497 


63.02 


8 


128.5 




140 


144 


16.509 


.627 


49.98 


9 


114. 4 




126 


130 


13.094 


.791 


39.63 


lO 


101.9 


108.0 


112 


116 


10,381 


1. 000 


31.43 


II 


90.7 


97.0 


lOI 


105 


8,234 


1.257 


24.93 


12 


80.8 


87.0 


91 


95 


6,529 


1.586 


19.77 


13 


72.0 


78.0 


82 


86 


5.178 


1.999 


15.68 


14 


64.1 


71.0 


75 


79 


4.107 


2.521 


12.43 


15 


57 I 


63.0 


67 


71 


3.257 


3.179 


9.86 


i6 


50.8 


55.0 


59 


63 


2,583 


4.009 


7.82 


17 


45.2 


49.0 


53 


57 


2,048 


5.055 


6.20 


i8 


40.3 


44-0 


48 


52 


1,624 


6.374 


4.92 


19 


35.9 


39.0 


43 


47 


1,288 


8.038 


390 


20 


32.0 


36.0 


40 


44 


1,022 


10.140 


3.09 


21 


28.5 


32.0 


36 


40 


810 


12.780 


2.45 


22 


25.3 


29.0 


33 


37 


642 


16.120 


1.94 


23 


22.6 


27.0 


31 


35 


509 


20 . 320 


1.54 


24 


20.1 


24.0 


28 


32 


404 


25.630 


1.22 


25 


17.9 


22.0 


26 


30 


320 


32.310 


.97 


26 


15.9 


20.0 


24 




254 


40.750 


.77 


27 


14.2 


18.0 


22 




202 


51.380 


.61 


28 


12.6 


17.0 


21 




160 


64.790 


.48 


29 


II 3 


15.0 


19 




127 


81.700 


.38 


30 


10.0 


14.0 


18 




100 


103.000 


.30 


31 


8.9 


12.5 






80 


129.900 


•24 


32 


8.0 


II. 5 






63 


163.800 


.19 


33 


7.1 


10.5 






50 


206.600 


.15 



Digitized by VjOOQ IC 



§ 13 DYNAMOS AND DYNAMO DESIGN 73 

An example of the use of this formula will be given later, 
in connection with the design of a dynamo. 

90. In Table II is given the properties of copper wire in 
sizes corresponding to those of the American or Brown 
& Sharpe gauge. This table gives in successive columns the 
gauge number; the diameter in mils of the wire when bare, 
single cotton-covered, double cotton-covered, and triple cot- 
ton-covered ; the area in circular mils; the resistance in ohms 
per 1,000 feet; and the weight in pounds per 1,000 feet. 
The omission of a size in the columns for the insulated 
diameters indicates that the size in question is not ordinarily 
insulated in that manner. Thus, larger wire than No. 10 
B. & S. is not ordinarily single cotton-covered, nor are 
smaller sizes than No. 25 triple cotton-covered, etc. 

91. In armatures, double and triple cotton-covered wires 
are chiefly used, although for very small wire, say, above 
25 B. & S., double silk or single cotton is preferred by some 
manufacturers, since double-cotton insulation occupies an 
excessive amount of room. Single insulation is not desir- 
able, unless the wire will have very little handling, for the 
covering may slip along the wire and leave a portion entirely 
bare. This is not liable to occur with double or triple 
coverings, since the direction of winding the insulation is 
different in the two layers. 

92. liStlmation of Hysteresis Ijoss. — The loss in the 
armature core, due to hysteresis, depends on the volume of 
the iron, the magnetic density, the frequency of the mag- 
netic reversals, and the quality of the iron; that is to say, 
each magnetic cycle requires a certain expenditure of energy 
per cubic inch, and two cycles will require twice as much as 
one. A magnetic cycle in a direct-current dynamo is accom- 
plished when the armature revolves through the angle of two 
poles; thus, in a bipolar there is one cycle per revolution, in a 
four-pole machine there are two per revolution, etc. The fre- 
quency \^ the number of complete cycles per second ; numeric- 
ally, it is the product of the number of pairs of poles and 
the revolutions per second. If a core of V cubic inches is 

44—11 



Digitized by VjOOQIC 



74 



DYNAMOS AND DYNAMO DESIGN 



§13 



subjected to a frequency of n cycles per secoad, and if at 
the particular magnetic density there is a loss of a watts per 
cubic inch and per cycle, the total loss due to hysteresis is 

W^^axVxn (17) 

93. Table III shows the loss in watts per cubic inch per 
cycle for an average quality of annealed iron, such as is ordi- 
narily used for armatures, and is here given for convenient 
reference in connection with the subject of dynamo design. 
Iron of very much poorer quality, that is, having a larger 
value for a, is undesirable, as the iron losses may so overheat 
the core as to decrease the capacity of the machine. For 
transformers and for generators with high frequencies, it is 
necessary that the value of a be as low as possible, and is 
usually about half of that given in the table. 

TABIiE in 
WATTS LOST PEB CUBIC IXCH PEB CTTCUB PBB SECOND 



Density. 

Lines 
Per Square Inch 


a = Watts Per 
Cubic Inch. 

I Cycle Per 
Second 


Density. 

Lines 
Per Square Inch 


a = Watts Per 
Cubic Inch. 

I Cycle Per 
Second 


30,000 
40,000 
50,000 
60, 000 
65,000 
70,000 
75,000 
80,000 
85,000 


.0042 
.0067 
.0095 
.012,8 
.0145 
.0164 
.0183 
.0202 
.0223 


90,000 
95,000 
100,000 
105,000 
110,000 
115,000 
120,000 
125,000 


.0244 
.0267 
.0289 
.0312 

.0337 
.0362 
.0387 
.0414 



94. Eddy-Current lioss. — The eddy-current losses are 
most difficult to compute, and, in fact, are practically impos- 
sible to predetermine with accuracy. This is especially the 
case if the slots are filed or milled after the armature punch- 
ings have been assembled ; such treatment connects the disks 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 76 

together and may greatly increase the eddy-current loss. 
Fortunately, they are usually of minor importance, and may 
be estimated with sufficient accuracy. A very good way is to 
assume the eddy-current loss to be proportional to the hyster- 
esis loss, which it very nearly is, and so increase the value 
computed for the latter by from 25 to 100 per cent. This 
will usually be found to be quite satisfactory unless other 
data is at hand on the subject. A very good rule is to allow 
75 per cent, over the values given by the tabic, for both 
hysteresis and eddy currents combined. This will usually 
be found to be very safe, but should, of course, be checked 
by tests with the first machine built, and the rule modified 
to suit the action of the type of dynamo in question. 

95. Radiating: Snrfitce and Temperature Bise. — The 

. sum of the PR and the hysteresis and eddy-current losses 
gives the total watts that must be radiated from the armature 
surface (exclusive of the commutator) ; if the total number 
of square inches of radiating surface is computed, the ratio 
of watts to the radiating surface gives the watts lost per 
square inch. By noting the number of watts that must be 
radiated per square inch, an idea can be formed as to the 
probable temperature rise. 

While it would seem as though these calculations should 
be simple enough, there are many ways of estimating the 
losses and also the radiating surface. One way is to esti- 
mate the total area of the armature, including the ends, 
and, if the core is ring-shaped, the inside cylindrical surface. 
Another way is to include only so much of the PR loss as 

' occurs in the slots, adding this to the iron loss, and taking 
for the radiating surface only the outside cylindrical surface 
of the armature core, omitting the area of the end connec- 
tions. Any method will^ of course, give satisfactory results, 
if it is obtained from tests of actual machines of approxi- 
mately the same type as those for which the heating calcu- 
lations are required. 

96. It is customary to limit the temperature rise of the 
armature to from 40° to 45** C. above that of the surrounding 



Digitized by VjOOQIC 



76 



DYNAMOS AND DYNAMO DESIGN § 13 



atmosphere. From numerous tests of drum-wound multi- 
polar armatures with both barrel and spiral-end connections, 
the following table of watts that can be radiated per square 
inch of armature core surface without the temperature rise 
exceeding 45° has been prepared. The calculations were 
made according to the second method just given. 

TABIiE IV 



Peripheral Velocity of 
Armature Core. 


Watts Loss Per Square Inch of Surface of 
Armature Core for a Rise of 45" C. 


Feet Per Minute 


Good Ventilation 


Poor Ventilation . 


2,ooo 
3,ooo 
4,ooo 
5,ooo 


2.4 

3.5 
4.5 
5-4 


1-75 
2.40 
2.80 
3.10 



By good ventilation is meant a ventilating duct in the 
armature core for about every 3 inches of length parallel to 
the shaft. These air ducts should preferably be wider than 
i inch to prevent their becoming clogged up with dust. By 
poor ventilation is meant narrow air ducts or none at all, 
windings covered by canvas or other material, or frames so 
shaped as to obstruct the free circulation of air. 

In the design of a dynamo to be given later, an example 
of the heating calculations is given. In designing, the 
computed watts lost per square inch of the armature-core 
surface should not exceed the value given in the table, 
because the machine would then probably not be capable of 
delivering its output continuously without overheating. 
Nor should the computed value be less than .8 of that in the 
table, since this would indicate that the design is too large 
to be economical of material, and should be made smaller by 
increasing the densities until the estimated temperature 
rise comes within the limits given. 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 77 

97. Heating calculations are perhaps the most unsatis- 
factory of any of those concerning dynamos. The iron of 
the armature core is liable to vary greatly in quality, and 
tests usually show very discordant results in samples taken 
from the same lot of iron. Again, the radiating qualities 
of a surface are somewhat uncertain, for a coat of varnish or 
a film of oil may make a difference of from 20 to 30 per cent. 
The foregoing rules must be taken, therefore, only as giving 
very approximate values, and data on such subjects as heat- 
ing and sparking must be collected for reference by each 
electrical designer for himself, from the factory tests of the 
generators as they are built. 



DESIGN OF THE FIELD MAGNET 

98. The field-magnet frame is designed especially to 
carry or conduct the magnetic flux from pole to pole. It 
must also be provided with the exciting coils, properly insu- 
lated and protected, and designed so as to set up the required 
magnetic flux when currents are passed through them. The 
materials of which the field magnet is made are selected, usu- 
ally, from considerations of economy, from one or more of the 
following metals: Cast iron, low-carbon cast steel, wrought- 
iron forgings, or sheet-iron or steel punchings. The mag- 
netic properties of these substances are shown in the curves, 
Fig. 41. These curves give the ampere-turns per inch of 
length of magnetic circuit for different values of the density. 
Now, it has already been shown that 

^ 3.192 X ampere-turns 

~ length of magnetic circuit in inches 

which may be written 

IT 

-J- is the ampere-turns divided by the length of the 

magnetic path in inches; hence, it is the ampere-turns 



Digitized by VjOOQ IC 



78 



DYNAMOS AND DYNAMO DESIGN 



§13 



per inch, and the value of this quantity is found by divi- 
ding the value of H by 3.192, or multiplying it by .313. 



i^nnn^ 




550 








500 








450 








A»7 








350 










/spooo 














1 — 




































































































— -- 




^ 


^ 






■■ 




























|_ 






























— i 






L 


- 
































































■^ 




^ 


- 


^ 


-^ 


L 










B 












































r— 


.^ 


- 




- 


=" 


fcOUW 
























nfl 


> 


^ 


ti 


/£ 


t^ 




■^ 


































c/ 


ej, 


f^ 


ii 












VH 






- 






,^ 


— ■ 




- 


- 
















^ 










/ 


f" 


J^ 


^ 


^ 






^ 


- 


^ 


— 


— 




"■ 












, 


^ 


y^ 




r 


i5f. 


; 


55 


^ 




r, 


^5 


//y 


1^ 






















% 






/ 


r 






^ 


^ 


Y?« 


7 


rr 


/A- 
































* 




/ 


' 




/ 


> 




W* 








































^ 90000 


J 


f 


/ 


^ 
















































I 


I 


V 

J 


/ 


















































v> eoooo 

4. 


y 


y 




















































^ 


// 






















































j» 


// 






















































^ ] 


V 








































^^ 


— 


- 


- 




•^ 


" 


^60000 -\ 


^ 
































^ 




"^ 


















\ "\ 


. 






















^ 
































- ^ 
















4 


y 


















































^ 


X 


















































/ 


/ 




















































A 


/ 




















































/ 






















































/ 


1 






































































































































































1 


' 
























































* 


























U 



























50 WO 150 ZOO 

Ampere -Turns per /nc/r Le/r^M. 
PIO. 41 



250 



XO 



99. From the curves, it is seen that cast iron is scarcely 
half as good as the others magnetically, but its cheapness in 
comparison and also its ability to be readily molded into 



Digitized by VjOOQIC 



§13 DYNAMOS AND DVNAMO DESIGN W 

any intricate shape sometimes outweighs its magnetic short- 
comings. Wrought-iron forgings are too expensive, unless 
of very simple shape, as, for instance, a straight, square, or 
round piece that may be cut from a long bar rolled for the 
purpose. Sometimes, on very small machines, drop for- 
gings of wrought iron or wrought steel are used. Steel cast- 
ings are very largely used, for they are comparatively 
cheaper than iron castings, since they are twice as good 
magnetically, although costing more per pound. Steel is 
cast at very high temperatures and intricate shapes are 
molded with more difficulty than with iron; also the shrink- 
age is much greater and the liability to form blowholes and 
shrink holes is greater on account of the higher tempera- 
tures of pouring. Where lightness is desired, as, for exam- 
ple, in the frames of railway motors, steel castings are used 
almost exclusively. 

100. Punch ings made from the iron or steel sheets may 
be readily assembled and riveted together into parts with 
rectangular sections with comparative economy, if the parts 
are small, for they do not require machining after being accu- 
rately punched; the saving in this way over castings helps 
to make up for the labor of building them up. Where the 
magnetic flux does not vary rapidly through a part made up 
of laminations, as in the magnet frame of a dynamo, the 
laminations may be riveted together; but this construction 
would be undesirable for an armature on account of the 
eddy currents through the rivets. Modern generators fre- 
quently have pole pieces of laminated iron to prevent pole- 
face eddy currents. 

MAGNETIC DENSITIES IN VARIOUS PARTS 

101 • Armature Core. — The ma^rnetlc densities used 
in the materials in modern dynamos depend not only on the 
material but also on the parts of the magnetic circuit for 
which it is to be used. The armature core is subject to 
hysteresis and eddy-current losses, which increase with the 
density, and therefore usually limit the density to such a 



Digitized by VjOOQ IC 



80 DYNAMOS AND DYNAMO DESIGN §13 

point that the losses will not be so great as to overheat the 
armature and lower its output. The density here is usually 
between 50,000 and 100,000 lines per square inch, the lower 
values being for machines in which the frequency of mag- 
netic reversals is high and the higher values for those in 
which the frequency is low. 

102. Armature Teeth*— In the teeth, the magnetic 
density should be quite high, say from 120,000 to 130,000 lines 
per square inch, as a high density here means a high reluc- 
tance, which will interfere beneficially with the cross-mag- 
netizing action of the armature. 

103. Pole Pieces and Magrnet Cores. — In the pole 
pieces, the density will generally be low, since the air-gap 
density is usually between 30,000 and 70,000 lines. It is, 
however, best to run the density 6f the magnet cores well 
up to the saturation for two reasons. In the first place, 
some part of the magnetic circuit should be saturated, since 
the machine will otherwise be unstable in its action. A 
small change in the ampere-turns will mean a comparatively 
large change in the flux and E. M. F., and it is best that the 
magnetic cores, rather than any other parts except the teeth, 
be saturated. Secondly, if the area and perimeter of the 
•magnet cores are made small, the weight of the copper in the 
exciting coils will be decreased. If the magnet cores are of 
4:ast iron, the density should be from 40,000 to 60,000 lines; 
while if of steel, laminated, or wrought iron, from 95,000 
to 105,000 is best. Cast iron is now seldom used for magnet 
cores of direct-current machines. 

104. Yoke. — The density in the yoke is usually not made 
very high, unless the type of frame is such that the path for 
the magnetism through the yoke is short. For cast-iron 
yokes, from 30,000 to 40,000 lines per square inch is used; 
while for steel and wrought-iron, the density should be 
70,000 to 90,000. Cast iron is very largely used for yokes 
on account of the ease of molding; besides, a fairly large 
cross-section of metal in the yoke is often desirable in order 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 81 

to give mechanical stiffness. With cast-steel yokes the 
cross-section necessary to carry the flux is sometimes not 
great enough to give mechanical stiffness. This point will 
be illustrated later on in connection with the design of a 
machine. 



GENERAIi FEATURES REIiATING TO MAGNET 
FRAMES 

105. The curves given in Fig. 41 are not to be consid- 
ered as absolute, for it is not infrequent to find materials 
whose properties differ by 10 per cent, or more from those 
given, especially in the case of steel castings. The compo- 
sition of the irons and their treatment have much to do with 
their magnetic qualities. In general, the purer the steel or 
wrought iron and the softer it is, the higher will be its mag- 
netic permeability. Chilling lowers the permeability and 
annealing raises it; also, impurities, such as carbon, that 
tend to harden the metal, usually lower the permeability, 
while those that tend to soften the metal raise its permea- 
bility when present in small quantities. 

Magnet frames are frequently constructed of more than 
one substance, for it is often desirable or necessary to have 
one part — such as the pole pieces — removable from the rest, 
so that the use of a different material involves no difficulties. 
Even where the parts are not required to be detachable, 
there may be one part, say of wrought iron or punchings, 
cast into another. Wrought-iron or laminated pole pieces 
are frequently cast into the yoke. 

106. The selection of the proper type of magnetic frame 
for a certain purpose is by no means a simple one, and the 
choice must be guided largely by experience. If the machine 
is not restricted, some type giving large radiating surfaces 
for the armature and field coils should be taken in order 
that the temperature rise from the losses shall not be exces- 
sive. If the machine is to be entirely or partially enclosed, 
the type of frame must be such as will readily lend itself to 



Digitized by VjOOQ IC 



82 DYNAMOS AND DYNAMO DESIGN g 13 

enclosing the parts of the machine. In street-railway 
motors, the casing is the magnet frame itself, which is of 
cast steel. Sometimes small bipolar machines are provided 
with covers made of brass, or some other non-magnetic 
metal ; but in larger machines it is possible to select a multi- 
polar type of such form that the casing does not form a path 
or circuit for magnetic leakage, so that it may be made of 
cast iron. 

107. Ma^rnetlc I^akagre. — Having selected the type of 
magnet frame and the material, or materials, of which it is 
to be made, it is a very simple matter to determine the areas 
of the various sections by dividing the total flux that must 
pass through the part by the permissible magnetic density 
for the material. In determining the flux, it is necessary 
to take account of magnetic leakage. The magnetic leakage 
is not easily determined, but accuracy in this case is not 
necessary, since the magnetic properties of the material 
vary so greatly that a considerable margin must be allowed 
anyway. The flux entering or leaving the armature under 
a pole is readily computed from formula 1, solved for 4^, and 
is a perfectly definite quantity. 

For purposes of calculation this flux is usually increased 
by a factor, called the coefflelent of magrnetic leakag^e, in 
order to obtain the flux in the pole pieces and yokes. The 
flux calculated from the formula is often termed the arma- 
ture tlnx, while that which is increased by including the 
leakage factor is called the field flax. The armature flux 
is used in the calculations of the densities in the air gap, the 
teeth, and in the armature core; the field flux is used in the 
calculations of the densities in the pole pieces, magnet cores, 
and yoke. Of course, the actual flux passing various points 
does not suddenly increase on account of leakage at the pole 
piece, but leakage increases it gradually as we pass up the 
pole pieces and magnet cores toward the yoke. However, 
to make the calculations simple this assumption is made, and 
in practice it is found that the results obtained from so 
doing are entirely satisfactory in most cases. 



Digitized by VjOOQ IC 



gl3 



DYNAMOS AND DYNAMO DESIGN 



83 



108* It is impossible to prevent magnetic leakage, and to 
compensate for it, it is only necessary to increase the sec- 
tional area of the field-magnet frame; so that where the 
leakage flux is only 10 or 20 per cent, of the useful armature 
flux, it is not a serious matter, but designs in which there is 
a leakage of from 25 to 50 per cent., or more, should be 
avoided, as the increase in weight of a corresponding amount 
for the magnet would involve an undesirable expense. 

The leakage coefficient is taken as unity plus the per cent, 
of leakage; thus a leakage coefficient of 1.2 means that the 
leakage flux, is 20 per cent, or .2 of the armature flux. It 
has been found by experiment that the value of the leakage 
coefficient does not vary greatly with the size of the machine, 
but it is practically constant for a given type, although it is 
somewhat smaller for the larger sizes. 



DETERMINATION OF AMPERE-TUBNS ON VIEUD 

109. As ha? already been stated, there must be a com- 
plete magnetic circuit provided for the magnetic lines, 
and in most types of field magnets 
there are several complete circuits, 
each one of which has linked with it, 
after the manner of Fig. 42, a field 
coil whose ampere-turns serve to set 
up the magnetic flux. In order to 
determine the proper strength of the 
field coil, or coils, in a particular cir- 
cuit, it is first necessary to map out 
the path of the magnetic lines through 
the circuit. To determine the length 
of the path for purposes of calcula- 
tion, it is necessary to determine the 
mean path, or the average between the shortest and longest 
paths. In the diagrams of field magnets shown in Figs. 31 
to 40, Part 1, the mean paths of the magnetic flux are shown 
by the dotted lines. 




Digitized by VjOOQ IC 



.84 DYNAMOS AND DYNAMO DESIGN §13 

110, Having computed the armature flux per pole from 
formula 1, and the field flux per pole by multiplying this by 
the leakage coefficient, the densities in the various parts 
may be determined if the areas of the sections are known, 
taking care to note and allow for the division of the flux, if 
any. Or, if the density in the various parts has been decided 
on, the area of any section is evidently the quotient of the 
flux through it divided by the density. Knowing the densi- 
ties, the ampere-turns required per inch of length of path 
may be determined from the curves. Fig. 41, for magnetic 
materials ; while for the air gaps, the ampere-turns are com- 
puted from the formula 

where Bg = magnetic density in lines per square inch; 
/ = length of path through air in inches; 
IT = ampere-turns required. 

The total ampere-turns for the magnetic circuit in ques- 
tion are now found by obtaining the product of the ampere- 
turns per inch by the length in inches of the mean path for 
each part of the circuit and then taking the sum of these 
products. An example of these calculations is given later 
with the design of a machine. 



FIEI.D WINDINGS 

111. Having determined the number of ampere-turns 
reljuired on a magnetic circuit, the number of turns is next 
to be calculated. Referring to the diagram of the magnetic 
circuit of the particular type of magnet frame under consid- 
eration, note the number of coils thereon, and if each is to 
be of the same strength, divide the total number of ampere- 
turns required by the number of coils on a single magnetic 
circuit; the quotient gives the number of ampere-turns per 
coil. In the radial-pole type, Fig. 40, Part 1, it will be 
noticed that there are two magnetic circuits threading 
through each coil, and that each magnetic circuit threads 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 85 

through, or links, with two coils. In this case there are two 
coils per magnetic circuit, and the fact that each coil sur- 
rounds two different circuits does not affect the calculations 
in any way, except that the magnet cores must be twice as 
heavy to conduct the flux of both circuits through the coil. 

113. Determination of Series-Field Winding:. — If 

the field coil is series-wound, the number of turns the coil 
should have is obviously the quotient of the number of 
ampere-turns required divided by the number of amperes 
developed at full load. The wire must have sufficient cross- 
section to safely carry the current without overheating. It 
has been found that the surface of a coil can radiate from 
^ to f watt per square inch of outside area, with a rise in 
temperature of about 40** C. To determine the proper size of 
wire for a series-wound coil, it is necessary to estimate the 
radiating area of the coil in square inches and take one-half 
or three-fourths of this as the permissible loss in watts W 
that can be safely dissipated. If / is the current in amperes 
in the coil and R is its resistance in ohms, then W ^ P R^ 

W 
or 7? = -j^. But, from formula 15, 

jp __ length in inches __ W 
"" circular mils ~" P 

Calling L the average length of a turn of wire and T the 
number of turns, then the total length in inches of the coil 
is Z, X Ty and if q is the area of the wire in circular mils, 



R = 



L T _ W_ 



L TP 
whence q = — Tjnr— (20) 

L X ITx I 
Formula 20 may be written q = -rp . But / T 

is the number of ampere-turns, so that we may write 



9 = 



mean length of a turn in inches X ampere-turns per coil X current 



watts dissipated per coil 



(81) 



Digitized by VjOOQIC 



86 DYNAMOS AND DYNAMO DESIGN §13 

However, the form given in formula 20 is as useful 
as any. 

113. Determination of Shunt Winding:.— If the field 

coil is shunt-wound, the size of wire is usually determined 
first, and the number of turns later in the following manner. 
Let e be the E. M. F. impressed at the terminals of a single 

shunt coil, then the shunt current i will be equal to -j^, whence 

e LT 

the ampere-turns iT ^ -jy x T^ but R = , which, sub- 

K q 

stituted in the above, gives 

'^n_ TxeXq _ exq 
TxL ~ L 

This may be written 

? = ^— (23) 



From formula 22 it will be noticed that the circular-mils 
area of a shunt wire is independent of the number of turns, 
the resistance of the coil, and the current flowing in it, for 
these qualities are not involved in the calculations. That is, 
the size of wire depends on the ampere-turns and not on the 
values of the current or turns taken separately. That this is 
so can be readily seen from an example. Suppose that a coil 
is subjected to a pressure of 25 volts, and that 10 amperes 
flows through, say, 300 turns, setting up 3,000 ampere- 
turns. If half the turns were removed, the resistance would 
be halved ; consequently, with the same E. M. F. at the ter- 
minals, the current would be increased to 20 amperes. The 
ampere-turns are / 7^ = 20 X 150 = 3,000, which is the same 
as before. Notice that in decreasing the number of turns 
the current is increased; this increases the number of watts 
lost in the coil and the heating of a shunt-wound coil there- 
fore depends on the number of turns of wire. 

To determine the number of turns of wire for a shunt coil, 
estimate the radiating area of the coil and determine how 



Digitized by VjOOQ IC 



§13 DYNAMOS AND DYNAMO DESIGN 87 

many watts W it can safely dissipate. The allowable current 

IV 
in the coil then is ^ = — and the correct number of turns is 
e 

^=T- = ^^ (23) 

114, Comparing the methods of calculating shunt- wind- 
ings and series-windings, it will be noticed that in series-coils 
the number of turns follows from the current and the ampere- 
turns, while the size of wire is determined from heating con- 
siderations. In shunt coils, the size of wire is determined 
by the ampere-turns, the mean length of a turn, and the 
E. M. F. impressed on the terminals, while the number of 
turns of wire is determined by the heating. 

115, Since the field coils surround parts of the magnet 
frame usually termed the mafirnet cores, it is evidently 
desirable that the cores have as smalk a perimeter as possible, 
so that the mean length of a field turn may be short, and 
thus the weight of field copper kept down, as the cost of this 
is quite a large item in the cost of a dynamo. The magnet 
cores must have sufficient area to carry the flux, and as the 
area required is much less for wrought iron, cast steel, or 
laminated iron or steel than for cast iron, magnet cores 
are rarely made of cast iron. Further, for a given area, a 
circle has the smallest perimeter of any figure ; and of rect- 
angular figures, the square has the least perimeter. There- 
fore, magnet cores are preferably made cylindrical, except 
in the case where laminated iron is used, when it becomes ^ 
difficult to make the section other than rectangular, and in 
this case the square is preferable. Where the departure from 
a circle or square is small, as in the case of an ellipse or a 
rectangle whose major and minor axes or sides do not differ 
greatly in length, the weight of copper required is not greatly 
affected; but designs where the major axis of an ellipse, 
say, is two or more times the minor should be avoided. 
The relation between the perimeters for the same areas in 
any cases may be readily computed from geometry. As an 
example, let it be required to compare a square and a circular 



Digitized by VjOOQ IC 



88 DYNAMOS AND DYNAMO DESIGN §18 

section. If we take the area of a circle 1 inch in diameter as 
a basis for comparison, then the perimeter or circumference 

O 1 A\ ft 

is 3.1416 inches. The area is -^~j — X d* = .7854 square 

inch. A square whose area is .7854 inch has a side 
>/. 7854 inch long = .8802 inch; therefore, its perimeter is 
4 X .8862 inch = 3.5448 inches, so that the ratio of the two 

. ^ . 3.5448 . .o • ^ u 

perimeters is oTTui' ^^ 1-13; in words, a square has a peri- 
meter 13 per cent, longer than a circle enclosing the same 
area. 

116. It will be noticed that formula 33, for the area of 
a shunt-field wire, involves the mean length of a turn, and if 
the perimeter is increased, the size of wire must be increased ; 
as the length of wire is thereby increased, the weight of wire 
is doubly increased by increasing the mean length of a turn, 
other things remaining constant. It might be stated that 
for the same number of watts lost in two shunt coils devel- 
oping the same number of ampere-turns, at the same termi- 
nal E. M. F., the weights of copper required varies as the 
square of the mean length of a turn. If the weight of wire 
required for the coils wound around a core of circular sec- 
tion, then, is 1, the weight required for the coils to surround 
a core of square section would be (1.13)' = 1.277, or 27.7 per 
cent. more. In this case, the difference in cost of the coils 
will be even more than this comparison would indicate, 
since the difficulty of winding a square coil increases the 
labor cost quite considerably. 

117. Space does not here permit a complete study of all 
the various types of field magnets and field-magnet windings 
that may be used, but the subject, fortunately, is one that can 
readily be studied from an inspection of machines, and also 
from the illustrated publications of manufact urers. The same 
is true of other parts, such as the designs of pedestals, bed- 
plates, rails, rocker-arms, brush holders, terminal boards, etc., 
and the student is urged to make such study of the construc- 
tion of generators as his opportunity and time may permit. 



Digitized by VjOOQ IC 



DYNAMOS AND DYNAMO DESIGN 

. (PART 3) 



DESIGN OF A 100-KILOWATT DYNAMO 



DESIGN OF ARMATURE CORE AND WINDING 

!• For the purpose of illustrating the principles of design, 
let it be required to work out a 100-kilowatt belt-driven 
dynamo. Armature windings are to be calculated for the 
three customary voltages, viz., 125, 250, and 500 volts, 
when the dynamo is running at a moderate speed, say 
between 600 and 600 revolutions per minute. 

!$• Estimation of Total Output. — We have the equation 

watts = - — —-. — -^ (1) 

10' X 60 ^ ' 

The watts in this equation are not the commercial watts 
output or the output at the machine terminals, but the total 
electrical output of the armature, which latter includes the 
electrical losses in addition to the commercial output. The 
electrical losses are but a small percentage of the total watts 
and may be neglected in preliminary calculations on sizes 
over 50 kilowatts, but in very small machines they would 
not be negligible. The percentage of electrical loss is, 
further, somewhat higher for the lower voltages. 

§14 

For notice of copyright, se« page immediately following the title pag*. 
44—12 



Digitized by VjOOQ IC 



DYNAMOS AND DYNAMO DESIGN 



§14 



The electrical efficiency U^ has been defined as the ratio 
of the net electrical output W to the total electrical out- 
put Wi^ or 

^ • _ W^ __ net electrical output 

* "" W^< "" net electrical output -|- electrical losses 

This may be written 

W 

Then, the total electrical output that is required from the 
armature under consideration is — jj — watts. 

3. The approximate values given in Table I may be 
taken for the value of U. for moderate-speed generators. 

TABIiE I 
ELECTRICAL. EFFICIENCIES OF DYNAMOS 



Output 


Electrical 


Output 


Electrical 


Kilowatts 


Efficiency 


Kilowatts 


Efficiency 


.1 


•75 


30 


•930 


•5 


.80 


50 


•950 


I.O 


.85 


100 


.960 


30 


.87 


300 


.970 


50 


.89 


500 


.975 


10 


.91 







4, It will be noted that formula 1 has the following 
unknown quantities: 

iL = percentage of armature surface covered by poles; 

D = diameter of armature in inches; 

L = length of armature in inches; 

Bg = average magnetic density in the air gap; 

K = number of ampere-conductors per inch of armature 

circumference; 
•S = speed of armature in revolutions per minute. 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 3 

There are in all six unknown quantities and but a single 
equation determining the relation between them; it follows, 
therefore, that this equation can determine but one of the 
six, and the others must be determined in some other way, 
or assumed. The problem of designing a dynamo varies 
very much according to which of these quantities are known 
or assumed. Each one of the six will be considered sep- 
arately, and the considerations ordinarily determining its 
value will be explained. 



CONBITIONS GOVERNING PRELIMINARY ASSUMPTIONS 

6. The percentage of armature surface covered by 
poles should be as large as possible, but some space must 
be left between the poles to provide a neutral region for 
commutation, as has already been explained. If the air- 
gap density is fixed, the greater the surface covered, the 
greater will be the total magnetic flux, and the output of the 
armature thereby increased; while with the total magnetic 
flux fixed, the greater the surface covered, the lower will be 
the magnetic density in the air gap and the less the number 
of ampere-turns required to maintain this density, so that the 
efficiency will be increased on account of the smaller mag- 
netizing currents required to excite the fields. In any case, 
it is an advantage to have the percentage large. It usually is 
between 60 per cent, and 80 per cent., the former value for 
small sizes, or for those designs in which the field coils are 
slipped over the end of the pole pieces. For the present 
design .75 will be assumed as the value of the percentage. 

6. The diameter of the armature is frequently the 
quantity that is determined by the equation. However, 
since it is involved to the second and sometimes to the third 
degree, the calculations are somewhat simplified if it is 
assumed or known. The diameter might well be determined 
in case the size of the machine is limited, as, for instance, in 
street-car motors. It is often desirable to make use of 
punchings from the center of larger armatures, and in this 
case, also, the diameter is limited. In designing a complete 



Digitized by VjOOQ IC 



4 DYNAMOS AND DYNAMO DESIGN §14 

line of machines, it is important that the sizes be properly 
related to one another, and it is best, therefore, to assume 
the various diameters to be used. For the problem under 
consideration, the diameter will be determined from the 
formula. 

7. The length of the armature core is not usually 
made more than 15 inches or 18 inches for drum windings, 
unless the speed and magnetic densities are low, because 
the E. M. F. developed by a single turn consisting of two- 
face conductors then becomes as high as the number of volts 
allowable between commutator segments. The inductance 
of the coils also increases with the length of the core, and 
where the coils lie in slots and each conductor carries a 
large number of amperes, say over 100, it becomes extremely 
difficult to prevent inherent sparking at the commutator 
with armature cores 18 inches or more long. Where the 
magnet cores are near the pole pieces, as is usually the case, 
it is desirable that the length parallel to the shaft be so 
taken as to make the pole faces suitable for round or square 
magnet cores, as these are most economical of field copper. 
In such cases, the length of the armature is related to its 
diameter and the number of poles. Suppose, for example, 
that for round magnet cores, it is desirable that the pole 
faces shall be approximately square, or, in other words, 
that the length of the armature parallel to the shaft shall 
be about equal to the length of the arc of a pole. This may 
be expressed in a formula ; thus, 

^=^'-f (3) 

where/ is the number of poles and the other symbols have 
their usual significance. It is convenient to substitute for L 
in formula 1 the value // D, where A is the ratio of the length 
of the armature to its diameter. For square pole faces, 
then, from formula 3, 

, _ Z __ nx^ 



Digitized by VjOOQIC 



§14 DYNAMOS AND DYNAMO DESIGN 5 

When the pole pieces are made of laminated iron, having 
projecting horns, the arc should be longer than the length 
of the armature if the magnet cores are to be square. The 
amount by which it should be longer depends on the ratio 
of the magnetic densities in the air gap and magnet cores 
and on the magnetic leakage coefficient. A good rule in 
such cases is to take the length of the armature core .6 or 
.7 the length of the arc for laminated pole pieces. Let it 
be assumed for the design at hand that the length of the 
armature is .65 times the length of the polar arc, and that 
the number of poles is 6. The value of h then becomes 

h = ^X .65 = ?:lill2L.l« X .66 = .266. say i 

That is, the length of the armature core parallel to the 
shaft will be \ the diameter of the core. This assumes that 
the design will have laminated poles. 

8. The difference between machines with and without 
laminated poles is quite marked. If the poles are solid, the 
length of the air gap must be greater and the slot opening 
narrower in order to prevent pole-face eddy currents, as has 
been already explained. The slots in this case are usually 
made with overhanging teeth, or are straight but narrow 
and numerous. On the other hand, machines with lami- 
nated poles may have very short air gaps and large open 
slots with several coils in each. These wide slots reduce the 
inductance considerably and, hence, tend to improve com- 
mutation. But this is offset by the fact that since there 
are many coils per slot, the armature will turn a considera- 
ble angle during the commutation of the coils in a single 
slot; the coils, therefore, will commutate in different parts 
of the neutral region between the poles, and they may not 
all commutate equally well. This will result in the blacken- 
ing or burning of one segment in the commutator for every 
slot. Such unequal sparking is very objectionable, since it 
causes the commutator to wear irregularly and become very 
rough. 



Digitized by VjOOQ IC 



6 DYNAMOS AND DYNAMO DESIGN §14 

9. Density In the Air Gap. — By density in the air 
gap is here meant the average magnetic density obtained by 
dividing the total armature flux per pole by the area of the 
armature core facing a pole. This quantity is usually taken 
as large as possible in order that the output of a machine 
may be large for its dimensions. It is limited by the mag- 
netic saturation of the teeth and by the relation between the 
sizes of the slots and teeth. In order that the air-gap 
density may be high, the percentage of the circumference 
occupied by slots should be small, but in this case, obviously, 
the ampere conductors per inch of circumference must also 
be small, and the output is decreased on this account, while 
with a large percentage occupied by slots, the teeth become 
so thin that the flux they will carry, as well as the air-gap 
density, is decreased, and the output is again decreased. 
Between these two extremes there is a maximum that it is 
desirable to attain. The average air-gap density is usually 
between 40,000 and 60,000 lines per square inch, but may be 
anywhere from 25,000 to 75,000 lines. The lower values are 
usual for smaller machines, such as fan motors, and the 
higher values for very large dynamos, usually with laminated 
poles. By this latter construction, the size of the slots may be 
greatly increased, so that their number is reduced, as is also 
the room on the circumference occupied by slot insulation, 
and, consequently, a little larger percentage of the armature 
circumference is left for the teeth. At the same time, lam- 
inated poles permit the use of short air gaps, so that higher 
magnetic densities may be secured without the necessity of 
too great a number of ampere-turns on the field to set up 
the flux across the gaps. For the present design, an air-gap 
density of 55,000 lines per square inch will be assumed. 

10. Ampere Conductors per Inch of Clrcumferenoe. 

As has already been stated, this is restricted by the room in 
the slots. It has been found that the losses due to resist- 
ance of the armature conductors are objectionably large if 
the conductors have an area of less than from 500 to 600 cir- 
cular mils per ampere of current flowing through them. 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN ? 

These losses, along with those incurred in the iron, namely, 
hysteresis and eddy currents, raise the temperature of the 
armature, but while this temperature may perhaps be kept 
within limits by especially good ventilation, as has been 
explained, the efficiency of the machine will be lowered to 
an undesirable degree. In very small machines, the circu- 
lar mils per ampere may be as small as 300, while in larger 
machines it may be as much as 800 or 1,000. Since there 
must be a certain amount of copper in the conductors per 
ampere, it is evident that the greater the number of amperes 
per inch of circumference of armature, the more room will be 
required in the slots. Now, the slots cannot be made indefi- 
nitely deep; first, because the inductance of the coils is 
increased, and, second, because the deepening of slots with 
parallel sides narrows down the width of the teeth at their 
roots and chokes the flux at this point. On this latter 
account, small diameters require shallow slots, and fewer 
ampere-conductors per inch can be accommodated than with 
larger diameters of armature. The ampere-conductors per 
inch usually vary from 300 to 700. On fan-motor sizes, it 
may be less than 100. For the machine to be designed 
later, 550 ampere-conductors will be assumed. 

11. It will be noticed that where the current density in 
the copper is high, that is to say, where the circular mils 
per ampere are low, there will be greater heating per pound 
of copper than where the current density is low. Now, in 
designs where there are many amperes per inch of circum- 
ference, there must be considerable copper per inch, and in 
order that the heating at the surface shall not be excessive, 
the current density in the copper must be lower than where 
the amperes per inch are less. That is to say, the copper 
per inch should increase more rapidly than the ampere- 
conductors in order to keep the heat liberated per square 
inch of armature surface at the same value. 

12. Speed in Revolutions per Minute. — The speed of 
electric generators that are direct-connected to engines or 
other prime movers must be designed to agree with that of 



Digitized by VjOOQ IC 



8 DYNAMOS AND DYNAMO DESIGN §14 

the driving agent, but dynamo speeds are not otherwise 
closely restricted. Where the machines are belt-driven, the 
speed of the belts should be from 4,000 to 5,000 feet per 
minute for machines with pulleys over 15 inches in diameter, 
and it has been found desirable to limit the peripheral speed 
of armatures to about the same values, although some classes 
of dynamos have peripheral speeds of 6,000 feet per minute, 
or even higher. 

13, Many very satisfactory moderate-speed machines 
run from 2,500 to 4,000 feet per minute, and 4,600 feet per 
minute is not an unusual value. Engine-type direct-con- 
nected machines, of course, have much lower peripheral 
velocities, especially if intended to be direct-connected to 
Corliss or other slow-speed type of engine. We will assume 
that the machine is to be belt-driven and the revolutions 
per minute are not restricted, so it will be best to restrict 
the peripheral speed to 4,000 feet per minute. If D is the 
diameter of the armature in inches^ and the peripheral 
velocity is 4,000 feet per minute, then the speed in revo- 
lutions per minute will be 5 = -^ p; — , because 4,000 

X 12 gives the number of inches traveled by the circum- 
ference per minute, and tt x jO is the circumference in 
inches. This value is to be put in the fundamental for- 
mula 1, in place of 5. 

14, It will thus be seen that the quantities to be assumed 
in the fundamental formula for the output may have widely 
differing values, depending on the size and type of machine. 
To lay down a set of rules for determining each quantity 
might be possible, but the information from which these 
rules would be derived is the private property of the various 
electric manufacturing companies and could not be obtained, 
therefore, for such a purpose. It is the skill of the electri- 
cal engineer in properly proportioning the quantities to each 
other that insures the success of a particular design, and a 
very wide experience is necessary before his judgment 
becomes reliable. 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN d 

16« General Dimensions of Armature. — Referring 
again to the fundamental formula, 

watts - ^-'D'L^^KS 
watts - j^. ^ gjj 

For each of these quantities except D, values as follows 
have been taken : 

watts = "".^^ = 12M00 = 104.166. or, say, 104,200 

a, = .u 

4 
B^ = 55,000 
K = 550 
^ __ 4,000 X 12 

•^ - VD 

Substituting these values gives 

104,200 = .75X^'X/^XjX 55,000 X 560 

4,000 X 12 1 

^ i:D W X 60 

whence, 

^ 104,200 X 4 X 10 ' XJ60 _ 

.75 X 3.1416 X 55,000 X 550 X 4,000 X 12 
D = \/ni = 27.04", or, say, 27" 

Z = ~ = ^" = 6i" 

e 4,000 X 12 4,000 X 12 

5 = ,2^ = 3.1416X27 = ^^^ ^-P"^- = ^^y^^^ 

16, Armature Conductors. — It is usually more difficult 
to wind generators for 500 volts than for 250 or 125 volts, 
especially the smaller sizes, and it is therefore best to try 
the 500-volt windings first. The total current output of the 

generator at 500 volts is — ' = 200 amperes. If the few 

amperes flowing through the shunt coils be neglected, as may 
be done, the total armature current may be taken as 



Digitized by VjOOQ IC 



10 DYNAMOS AND DYNAMO DESIGN §14 

200 amperes. Now, taking a single series-winding for trial, 
there will be but two paths, and 100 amperes will be the cur- 
rent per path or per conductor. With 560 ampere-con- 
ductors per inch and 100 amperes per conductor, the total 
number of face conductors will be 

550 X ^ i^ _ 550 X 3.1416 X 27 _ 
100 "■ 100 

because the number of inches in the periphery is w /?, and 
550 y.'K D will give the total volume of current in all the 
conductors. This divided by the current per conductor 
evidently gives the number of conductors. 

17. Minimum Kamber of Commutator Segrments. 

The average volts per commutator segment should not be 
more than about 15, or there will be danger of the machine 
flashing over from positive to negative brushes. The least 
number of segments between adjacent brush points should 
then be ^-^ = 33.333, and since there are six poles, the total 
number of segments should not be less than 6 X 33. 333 = 200. 
With 466 face conductors arranged into a winding with one 
turn per coil, or two face conductors, there would be 233 coils 
and segments, which number is greater than 200, and is 
therefore safe. 

For 250 volts, the average volts per segment should not 
be greater than 10, and for 125 volts not greater than 7. 
There is not much liability of these lower voltages flashing 
over, but other requirements fix the above limits. For 
125-volt machines, for example, with a modern commutator 
arranged for carbon brushes, if the volts per segment were 
made 10, there would be but 12 segments from brush point 
to brush point, and each of these would then be very thick 
and would not make a good running commutator. These 
values are all taken from practice and are known to be satis- 
factory. Of course, they must all be exceeded in very small 
machines, for the segments should not be much less than 
•j^ inch thick in any case, because the percentage of mica in 
the commutator then becomes so great that it is liable to 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 11 

wear unevenly and prove unsatisfactory. The product of 
the inductance of the coils and the current in each con- 
ductor must be kept low in order to commutate the currents 
successfully, and in machines of very large current outputs, 
this may necessitate the use of very many more segments 
than the limiting values just given would indicate. At best, 
the foregoing method of approximating the number of seg- 
ments is but a rough guide, and a complete investigation of 
a particular winding for inductance and sparking is too com- 
plicated to be given here. 

18, Number of Slots and Coils. — For a single series- 
winding, the number of commutator segments and the 
number of coils must fit the formula 

where C is the number of segments, and y is any whole 
number. In this case, 

C = I X 78 - 1 = 233 

therefore 233 coils will wind all right. However, 233 is a 
prime number, and it would therefore be necessary to have 
233 slots to accommodate the winding. The expense of 
punching the disks and of winding the coils increases with 
the number of slots, and it would be better to wind 2, 4, 
or 5 coils in each slot. Three or 6 coils per slot being 

divisible by ^, or 3, makes the total number of coils divisible 
by =^, and will not wind properly into a single series- winding. 

At 

Four coils per slot will answer very well. The student will 
understand the arrangement of the 4 coils by referring 
ahead to Fig. 18. The 4 coils are taped together to form 
a single element that occupies the top and bottom halves 
of two slots separated by the distance between centers of 
poles. The number of coils of one turn each is therefore 
four times the number of slots. The nearest number to 233 



Digitized by VjOOQ IC 



12 



DYNAMOS AND DYNAMO DESIGN 



§14 



divisible by 4 is 232, and the number of slots required is 58. 
We have, then, 

C = I J ± 1 = I X 77 + 1 = 232 • 

This number will therefore wind. If too few slots are 
used, the machine will be liable to hum when running. The 
coils should each span about one-sixth of a circumference, 
or 10 slots, since 10 is the nearest whole number to y. 

19. Armature Conductor. — The size of the conductor 
required to carry 100 amperes may be determined by allow- 
ing about 600 circular mils per ampere. A copper strip | inch 
wide and .075 inch thick with rounded edges, as shown in 
Fig. 1, has an area of 68,125 circular 
mils, or 681 circular mils per ampere 
of current. This area consists of a 
rectangle .076 in. X .650 in. and 
two semicircles .076 inch in diameter. 
The rectangle has an area of 76 X 660 = 41,250 square 
41,250 




:075 



FlO. 1 



mils, or 



= 62,500 circular mils, approximately. The 



.7854 

two semicircles combined form a circle .075 inch in diam- 
eter; therefore, their area is 76 X 76 = 6,625 circular mils. 
The total area of the conductor, then, is 52,500 -f- 6,626 
= 58,125 circular mils. 

BO. Insulation of Armature Conductors. — To insu- 
late the conductors from each other, alternate conductors 
should be wrapped with cotton tape, half 
lapped, and the four conductors that are 
to go into one slot should then be taped 
together. The tape, which is usually about 
.009 inch thick by f inch wide, is double 
thick on a side on account of the lapping, 
or .018 inch on a side, so that each wrap- 
ping adds .036 inch to the thickness. The 
width of the four coils complete will be about 
as follows: There are two wrappings of tape 
on the conductors and one over all, or three wrappings of 




Pio. % 



Digitized by VjOOQ IC 



814 DYNAMOS AND DYNAMO DESIGN 13 

.036 inch make a total for insulation of .108 inch; the 
thickness of the four copper bars is .075 inch x 4 = .3 inch, 
making a total of .408 inch. Allowing .020 inch clearance, 
or extra room, makes the total .428 inch, which may be 
taken as .43 inch. In the depth there are but two wrap- 
pings over the conductors, so the total thickness of insula- 
tion is .072 inch; the copper itself is .625 inch deep, making 
a total of .697 inch, or .7 inch. See Fig. 2. 

21, Slot Insulation. — The slots should be insulated by 
a cell made up of a sandwich, as it is called, of two layers 
of presspaper each .015 inch thick, with a thickness of oiled 
muslin .010 inch thick between. Oiled muslin isa good quality 
of muslin dipped in linseed oil and dried. Its thickness varies 
from .007 inch to .015 inch, depending on the quality of the 
muslin. The thickness of the cell would be about .040 inch, 
which should withstand 2,000 volts or more without break- 
ing down or being punctured. The cell is first put into 
place for winding; then the bottom coil is inserted, and over 
it a strip of the same sandwich is 
placed to separate the upper and 
lower coils in the same slot. After 
the upper coil is in place, the cell 
is lapped over it at the top and a 
wooden or fiber wedge driven into 
the nicks at the top of the slots 
to retain the windings against the 
centrifugal force while running. 
This construction avoids the use 
of binding wires over the armature 
core; these bands are somewhat 
liable to be broken, especially 
where small air gaps are used. 

23. Dimensions of Slot. — The 

width of the slots is the sum of two 

thicknesses of the cell, one on each 

side, and of the coil, or 2 X .040 ^'®* ' 

+ .43 = .51 inch, a little over ^ inch. For the total depth. 



Digitized by VjOOQIC 



14 DYNAMOS AND DYNAMO DESIGN §14 

there must be added to the depth of two coils, four thick- 
nesses of the cell and the thickness of the hard-wood retaia- 
ing wedge or (2 X .7) + (4 X .040) + fV = 1''475 inches, 
or 1.75 inches. Fig. 3 shows the shape of the slot and the 
arrangement of the conductors. 

23. Calculation of Total Flux. — From the equation 

£ = ;r^- — T^ X — , the total flux from or to one pole is 
60 X 10" m 

^ £ X 10" X 60 X w . w u T^ ' ^u ^ . i Ti^ \jr u^ 
* = „ ^ , m which £ is the total E. M. F. 

developed in the armature. In the present design, assume, 
say, 15 volts as the loss in volts due to the resistance of the 
armature winding, commutator, brushes, series-field, and all 
connections; then, since the voltage at the terminals is to be 
500, the value of E in the above equation is 515 volts. It 
has been decided that w, the number of paths through the 
armature winding, is 2; Z, the total number of face con- 
ductors, is 464; /, the number of poles, is 6; and 5, the 

, . • 4. • ««.« XT ^ 515 X 10* X 60X2 
revolutions per minute, is 575. Hence, ^ = . — -^r— • 

= 3,860,000 magnetic lines of force. 

34. Density In Teeth. — There are 58 teeth in all for 
six poles, so the number per pole is ^; but, since the poles 
cover only 75 per cent, of the armature, the average num- 
ber of teeth opposite a pole is ^ x . 75 = 7^^. So the flux 
of 3,860,000 lines passes through an average of 7 J teeth. 
The density at the tooth roots is an important point and 
should be determined. The diameter of the armature is 
27 inches and the slots are 1 J inches deep, so the diameter 
at the roots is 23J inches. The distance from center to cen- 

^ w . ^u . • 23.5 X 3.1416 , o«-o • u 
ter of slots at the roots is — = 1.273 inches. 

Do 
The slots are .51 inch wide, so the width of a tooth at the 
roots is 1.273 - .51 = .763 inch. See Fig. 4. 

The armature core is to be 6 J inches long, parallel to the 
shaft, and in the middle of this there should be a ventilating 
duct say J inch wide, leaving 6^ inches as the length of the 



Digitized by VjOOQ IC 



§14 



DYNAMOS AND DYNAMO DESIGN 



15 



punchings. Now, in order to prevent eddy currents from 
disk to disk, they should be insulated from each other, and 



■lrt« 




l-r^i^- 



FIO.4 

a very good way of doing this is to coat them with japan or 
put punchings of thin paper between them. The iron 
punchings are usually about .015 inch thick, so that even 
the thinnest insulation occupies a considerable percentage of 
the room. As about 85 per cent, of the punchings may be 
taken as iron, the net length of iron in the armature core is 
.85 X 6J = 5.31 inches. 

The area of iron at the roots of each tooth, then, is 
5.31 X .763 = 4.05 square inches. The area of 7^ teeth is 
then 29.4 square inches. Through this area the total flux 
of 3,860,000 lines pass, so the density at the tooth roots is 

q Qf^A 000 

' ' — = 131,300 lines per square inch. This density is 

about right; it should be under 140,000 lines and preferably 
over 120,000 lines. A high density at this point is of impor- 
tance in preventing, or tending to prevent, the effects of 
armature reaction. 

When the magnetic density in the teeth is higher than 
100,000 lines per square inch, quite an appreciable part of 
the magnetic flux will pass up through the slots and other 
non-iron paths between the teeth, and in the above compu- 
tations it is assumed that all of the magnetic flux passes 



Digitized by VjOOQ IC 



16 DYNAMOS AND DYNAMO DESIGN §14 

through the iron. To correct the tooth-root densities by 
allowing for the flux that passes up through the slots is 
ordinarily a needless calculation, for the error involved by 
computing in the foregoing manner is not of sufficient 
amount to be counted. 

25. Inside Diameter of Core. — The magnetic flux from 
each pole divides in the armature core below the teeth, half 



•i 



Fio.5 

going to each side (see Fig. 6), and, therefore, the flux 
through the armature core below the teeth is ^aa^JLIUi 
= 1,930,000 lines. The magnetic density in the armature 
core is usually from 50,000 to 100,000 lines per square inch. 
If the lower value is used, the amount of iron required will 
of course be greater, but the losses due to hysteresis will be 
much less, while the reverse is true for the higher densities. 
Say a density of about 60,000 lines is assumed for the present 
design; then, the area of the core below the teeth should be 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 17 

^l^gjgg = 32.2 square inches, approximately. Now, the net 

length of iron in the core has been found to be 5.31 inches, 

so the radial depth of the iron in the core below the slots 

32 2 
should be ■— ^ = 6.06 inches. We will make this depth, 
6.31 

say, 6^^ inches, in order to have even dimensions. The area 

of cross-section of the iron in the path in the armature core 

will then be 6.25 X 5.31 = 33.2 square inches. The radius 

of the bottom of the slots is 11| inches, and deducting 

6:J^ inches from this leaves 5^ inches for the radius of the 

central hole in the punchings, or 11 inches for the diameter. 

These dimensions are indicated in Fig. 5, which also shows 

some of the field dimensions calculated later on. 



HEATIXG CALCUXATIONS 

26, Hysteresis Ijoss. — As has been explained, there are 
two limits to the output of a dynamo, the temperature rise, 
or heating limit, and the sparking limit. To determine 
whether a design approaches these limits is somewhat diffi- 
cult, especially in regard to sparking. To calculate the 
probable rise in temperature, it is necessary to compute the 
losses per square inch of surface in the armature, as has 
already been explained. These losses consist of hysteresis 
and eddy-current losses in the iron of the armature core, 
and of PR losses in the conductors. In a six-pole dynamo 
there will be 3 cycles per revolution, or at 575 revolutions 
per minute, Yt^ X 3 = 28.75 cycles per secpnd. The density 
at the tooth roots at full load is 131,300 lines per square 
inch, and the width of a tooth at this point is .763 inch. At 
the tops of the teeth the width is .95 inch, so, since the 
same flux goes through the top as through the roots, the 

763 
density at the tops is 131,300 X —^^ = 105,500 lines per 

square inch, nearly. The average density in the teeth, then, 
is the average between 131,300 and 105,500, or 118,400 lines 
per square inch. The density in the core beneath the teeth 
is about 60,000. 

44—18 



Digitized by VjOOQ IC 



18 DYNAMOS AND DYNAMO DESIGN §14 

The width of a tooth at the top is .95 inch, and at the 
bottom is .763 inch; so the average width is .856 inch. The 
depth of a tooth is 1.75 inches, and the net length of iron is 
6.81 inches; so the volume of 58 teeth is 58 X .856 X 5.31 
X 1.75 = 461 cubic inches. The volume of core below the 

teeth is J X (23.5' — 11') X 5.31 inches = 1,798, or, say, 

1,800 cubic inches. From the table of hysteresis losses, it 
is found that for a density of 115,000, the loss per cubic 
inch per cycle per second is .0362 watt, while for 120,000 it 
is .0387 watt; hence, for 118,400 it would be about .0379 watt 
per cubic inch per cycle per second. The hysteresis loss 
in the teeth is, then, 461 X .0379 X 28.75 = 502 watts. From 
the table, the loss for 60,000 lines density is .0128 watt, so 
the hysteresis loss in the core proper is .0128 X 1,800 X 28.75 
= 663 watts, approximately. The total loss due to hystere- 
sis, then, is 1,165 watts. 

27. Eddy-Current lioas. — It is impractical to compute 
the eddy-current losses, as has been explained, but they may 
be estimated approximately as proportional to the hysteresis 
loss. Suppose it is assumed that with the care the lam- 
inations are insulated in the present design, the loss is 
three- fourths as large as the hysteresis loss, or Jx 1,165 
= 875 watts, approximately. When the machine is built, 
this should be checked up with the tests and the assumed 
factor corrected in future similar designs. The total iron 
losses then become 2,040 watts, or a little more than 2 per 
cent, of the output. 

38. y R lioss In Slots.— The heat developed by the P R 

losses in that part of the armature winding that is in the 

slots may be calculated as follows: The resistance of a single 

- J . u • D length in inches 6.75 

face conductor when warm is /c = — :^ — -. rj — = -rTrTzr?- 

circular mils 58,125 

In each conductor is 100 amperes, and there are 464 face 

conductors in all, so the total loss is 464 X 100* X ^q\^^ 

oo,l«o 

cs 589 watts. It must not be thought that this is all of the 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 19 

armature PR loss; it is only that part which is developed 
within the slots, and the part developed within the end 
connections has yet to be considered. The total watts 
developed in the core and slots is 2,040 + 539 = 2,579. 
Now, this heat has to be radiated from the surface of the 
armature core and of the ventilating flues, etc. The outside 
cylindrical surface of the core \s it D L = 3.1416 X 27 X 6.75 
= 573 square inches. Each square inch of this surface must 
then radiate ^^ = 4.6 watts. 

From Table IV, Part 2, it . is seen that, tor a peripheral 
speed of 4,000 feet per minute, which is the speed of this 
armature, 4.5 watts per square inch of core surface can be 
radiated, and in the design there is just 4.5 watts developed 
per square inch. The ventilation must therefore be made 
good ; for, otherwise, the dynamo will run too hot, as there 
is no margin. 



WINDING FOR »50 VOIiTg 

29. Thus far the 500-volt armature winding alone has 
been considered. For 250 volts, the number of face con- 
ductors in series on a path between brushes should be 
halved. Half of 232, or 116 segments, would be too few for 
a 250-volt commutator, because the average volts per bar 
would be too high, and it will therefore be necessary to 
use a double series-winding. This winding requires that 

the number of coils C = ^ X J ± 2, or 232 = | X 78 — 2. 

The pitch of the winding, then, is 78; whereas before for 
the single-series winding it was 77. As has been already 
explained, the change from a single to a double series- 
winding changes the number of paths from two to four, and 
since the total number of conductors on all paths remains 
the same, the number on each must be halved, as is the 
voltage also. The number of paths being doubled, the 
current-parrying capacity of the winding is also doubled, 
and since the voltage is halved, the output is not changed. 



Digitized by VjOOQ IC 



DYNAMOS AND DYNAMO DESIGN §14 



VnHTDTSG FOR 1«5 VOLTS 

30, Style of Winding:. — For 125 volts, a quadruple 
winding would give the correct voltage, but it could hardly 
be kept from sparking. . In fact, the double series- winding 
used for 250 volts is objectionable for high speeds and also 
for high voltages, so the quadruple winding is out of the 
question. A single series-winding for this voltage would have 
one-fourth as many segments as for 500 volts, or 58, which 
would give an average voltage of 12.9 per segment, and it 
has been stated that this last should not exceed 7 volts. 
Therefore, this winding should have at least ^^^ X 6 
= 107 segments. A double series-winding with 2 X 58 
= 116 segments would do, but a single parallel winding 
with 3 X 58 = 174 segments would be far superior as regards 
freedom from sparking. A parallel winding has as many 
paths as poles, or in this case six, while a single series-wind- 
ing has but two, so a single parallel winding, therefore, 
will require three times as many segments as a single series- 
winding for the same voltage. 

31. Cross-Connecting: Rlngrs. — As has been explained, 
a parallel winding is liable to peculiar commutator phenom- 
enon called bucking, and to prevent this, the winding should 
be cross-connected in the manner previously described. In 
the present design, four cross-connecting rings are about as 
few as could be relied on to prevent bucking. These rings 
should be connected to points that always remain at the 
same potential, or to points two poles apart. Therefore, 
the rings should be connected as follows: Ring No. 1 to 
segments 1, 59, and 117; ring No. 2 to segments 16, 74, 
and 132; ring No. 3 to segments 30, 88, and 146; ring No. 4 
to segments 45, 103, and 161. 

33. Size and Arrangrement of Conductor. — Using 
the same number and size of slots as before, there would 
be three coils per slot; 100 kilowatts at 125 volts is equiva- 
lent to 800 amperes, so each of the six paths should carry 



Digitized by VjOOQ IC 




§14 DYNAMOS AND DYNAMO bESiG^f H 

133 amperes. A conductor .1 in. x | in. [see {a), Fig. 6] 
has an area as follows: 100* + 625 X 100 

4 
X - = 76,845 circular mils, or 577 circu- 

lar mils per ampere. To insulate the 
conductors, the outside bars are taped, 
leaving the middle one bare; then the 
three are taped together, as shown at 
(*), Fig. 6. The width of the three 
conductors is .3 inch, and three wrap- 
pings of tape at .036 inch makes 
.108 inch, which leaves .022 inch clear- 
ance to bring the coil up to .43 inch 
wide, the width of the former coils. ^j 

See Fig. 8. f,o. t 



DESIGN OF COMMUTATOR 

33. Diameter and Peripheral Speed. — The commutator 
on modern machines has ordinarily a diameter between .6 
and .8 of the diameter of the armature. In this design the 
commutator will be made, say, 19 inches in diameter, or 
about .7 of the armature diameter. The peripheral speed of 
the commutator is best kept below 2,500 feet per minute, 
although many run as high as 3,500 or 4,000 feet per minute; 
but if a commutator gets out of true when running at a high 
peripheral speed, it will rapidly grow worse. The peripheral 
speed of a 19-inch diameter surface running at 575 revolu- 
tions per minute is 2,860 feet per minute. 

34. Thickness of Brushes. — With poles covering 75 per 
cent, or more of the armature surface, the neutral region on 
the commutator surface is usually so narrow that wide brushes 
short-circuit active coils and thus cause the commutator 
and armature to heat unnecessarily. Since a given distance 
around on the commutator means a greater difference of the 
potential on a high-voltage than on a low-voltage machine, it 
follows that the brushes should be thinner for 500 volts than 
for 125 or 250 volts. Just what thickness of brush can be 



Digitized by VjOOQ IC 



22 DYNAMOS AND DYNAMO DESIGN §14 

used in any given case is usually determined by trial after 
the machine is built; but for purposes of calculation, it may 
be assumed, from the known results in similar designs, that 
for the 500-volt windings ^^-inch brushes may be used; 
|-inch for 250 volts, and f-inch or |-inch for 125 volts. 
There is no reliable rule giving the number of segments the 
brushes should span, and such rules are not even approxi- 
mate, as can be seen by applying them to the present design. 

35. Brusli-Contact Area. — For 500 volts, the current 

, , - . watts 100,000 ,.^^ ^, . 

developed is — j— = — = 200 amperes. This current 

goes from the negative brushes into the commutator, and 
from the commutator to the positive brushes, so that the 
200 amperes is carried by three negative brush points, and at 
each brush point 67 amperes must pass either in or out of the 
commutator. Now, the current density at the contact sur- 
face for carbon brushes should not be more than 40 amperes 
per square inch, 30 to 35 amperes being good values. For 
copper brushes, the density may be as high as 100 amperes 
per square inch, and it is frequently higher than this. Cop- 
per brushes are little used on modern direct-current machines. 
At 30 amperes, each brush point would require about 
2.2 square inches area of contact surface. Making the 
brushes say j\ inch thick by If inches wide, each would have 
an area of .765 square inch, or three of them on each brush 
stud would have a combined area of 2.3 square inches, which 
is about the required amount. 

For 250 volts the current output is 400 amperes, so 
4.4 square inches contact area are required per brush 
point. Using |" x If" brushes, each would have an area of 
1.09 square inches, so that four of them would be required 
at each brush point. 

For 125 volts, in the same way, 8.8 square inches per 
brush point are required, which is obtained by using •}" 
X 1|" brushes, six at each brush point. 

This completes the electrical design of the armature and 
the commutators. The mechanical design and construction 
of these will be taken up later. 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 23 



THE MAGNETIC CIRCUIT 

36. Pole Pieces. — The pole pieces will be built up of 
punchings about ^ inch thick, riveted together between 
cast-steel, malleable-iron, or wrought-iron end plates | inch 
thick. As there are six poles covering 75 per cent, of the 

armature, each must include an angle of .75 X —^— = 45°. 

6 

37. Flux in Pole Pieces. — The flux to or from each 
pole has been found to be 3,860,000 lines. Now, with similar 
designs it has been found that the coefficient of magnetic 
leakage varies from 1.1 in very large machines with short 
air gaps to 1.2 in very small machines with comparatively 
long air gaps. Assuming 1.14 for the leakage factor in the 
present design, the total magnetic flux in the pole pieces 
becomes 1.14 X 3,860,000, or 4,400,000, approximately. 

38. Magrnet Cores. — The density in the magnet cores 
should be from 85,000 to 95,000 lines per square inch. It is 
possible to use as high as 100,000, or even more, but this 
requires that only iron of the best magnetic qualities be used, 
otherwise the ampere-turns required to force the flux 
through the iron will be excessive. Both iron and steel, as 
at present manufactured, are quite irregular in magnetic 
qualities, and a low density usually will save the expense of 
having to reject parts made up of iron of which the poor 
quality was not suspected until finished. 

At 95,000 lines per square inch, the area of the magnet 
cores for 4,400,000 lines should be 46.32 square inches. 
These cores will be about 95 per cent. iron. The percentage 
of iron is greater than on the armature, because the punch- 
ings are thicker and no insulation is required between them. 

The gross area of the cores, then, should be — ^— = 48.76, or, 

say, 7 inches square, or 49 square inches. The pole pieces then 
are J inch longer than the armature core, which was 6^ inches. 
The steel end plates on the poles would have eddy currents 
induced in them if they extended to the pole face, so it is 



Digitized by VjOOQ IC 



u 



DYNAMOS AND DYNAMO DESIGN 



§14 



customary to cut them short | inch or more. In this case, 
then, the pole face will be only 6J inches long, while the 
armature is 6J inches. 

The pole-piece punchings are 7 inches wide at the magnet 
cores, while over the pole tips, or horns, as they are called, 
the punchings are about lOf inches wide (see Fig. 7). The 
punchings are made with shoulders at a to support the field 
coils. The dimension d should be such that the magnetic 




Oap 



Pio. 7. 
density at the contracted point will be about 130,000 lines 
per square inch. The tips, or horns, project about If inches 
on each side from the body of the pole piece, and the length 
of the arc of the pole face is about 10| inches. Now, from 
this pole face, 3,860,000 lines pass to or from the armature; 

therefore, about -^ X 3,860,000 = 636, 000 lines pass from the 

tips. Since there is considerable leakage from pole tip to 

pole tip, say the flux through d is taken as 1.14 X 636,000 

= 725,000 lines. At a density of 130,000, the area of d 

should be 5.58 square inches. The net length of iron in the 

poles is .95 X 7 = 6.65 inches; so the dimension 6 in Fig. 7 

5 58 
should be ^-— = .839 inch, or say, | inch. Iii order to 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 26 

estimate the ampere-turns required for the magnetic circuit, 
it is necessary to assume some dimension for the radial 
length of the magnet cores. The magnet cores are usually 
from 1 to 1.5 times as long as they are wide, so as a basis for 
further calculations we shall take the dimensions shown in 
Fig. 6, making the inside diameter of the yoke 45^ inches 
and the bore of the surfaces where the poles fit on, 44} inches. 
These dimensions can, if found necessary, be modified later 
after the space for the field coils has been accurately 
determined. 

39. Air Gap. — In order to get a good magnetic fringe 
for commutation, and also in order to prevent humming, 
the air gap at the pole tips is usually made considerably 
greater than in the middle of the pole. In the case where 
solid poles are used, the same effect is accomplished by nosing 
the pole tips, as it is called, i. e., making them pointed, so 
that a tooth enters under the poles gradually. The nosing 
is usually not greater than is necessary to allow one tooth 
to pass completely under the pole while the next one is just 
under the tip of the nose. In the present problem, the air 
gap is assumed to be ^ inch, except at the tips, where it 
is I inch. The bore of the poles, then, is 27} inches, and 
it is further assumed that the air gap increases by a tangent 
line from the }-inch gap to the |-inch gap. See Fig. 7. 

40. Yoke. — Assuming that the yoke is made of cast 
steel, it will be quite safe to work it at a density of about 
75,000 lines per square inch. If it is certain that the quality 
of the steel will always be good, a density of 90,000 or even 
higher would be better, since less material would be required, 
but the magnetic quality of steel castings varies consider- 
ably, and since it cannot be readily determined whether a 
particular casting is good or bad until it is machined, a con- 
siderable loss is incurred if castings must occasionally be 
rejected; for this reason the lower density is preferable. 
The length of the magnetic path in the yoke is considerable, 
and if the iron requires many ampere-turns per inch to set 
up the flux, a considerable part of the available ampere-turns 



Digitized by VjOOQ IC 



26 DYNAMOS AND DYNAMO DESIGN § 14 

of the field coil will be required at this point ; the remainder 
may not be sufficient to set up the flux in the rest of the 
magnetic circuit, so the total flux will be too small. The 
voltage, therefore, will be too low, and unless the speed 
is increased, the windings or the castings will have to be 
changed. 

The flux from each pole divides at the yoke, half going 
right and half going left, so that the yoke section must 
carry half the flux in the poles, or about 2,200,000 lines. At 
75,000 lines per square inch, the area required for the sec- 
tion of the yoke is thus 29.3, or, say, 30 square inches. In 
order that the yoke shall be strong and stiff and at the same 
time have a heavy appearance, which is considered desirable 
in modern machinery, it will be made channel-shaped in 
cross-section, as shown in Fig. 5. 

41. A preliminary sketch of the field should now be 
made, as shown in Fig. 5, and the general dimensions of the 
machine may be thus determined. The mean path of the 
flux is shown by the dotted lines. These are lines that 
divide the middle of the section of the magnetic path. From 
the sketch, the length of the mean paths in the various parts 
can be measured off by means of dividers. 

42. It was assumed that the average air-gap density 
should be 55,000 lines per square inch, and while this aver- 
age air-gap density is about correct, still it is of no use in 



Pio. 8 



determining the number of ampere-turns required to set up 
the flux across the air gap. In order to obtain actual air-gap 



Digitized by VjOOQIC 



§14 



DYNAMOS AND DYNAMO DESIGN 



27 



densities, the paths or courses of the lines across the gap 
must be known, and since this is somewhat complicated, 
approximations are made for the purpose of simplifying the 

calculations. In Fig. 8, where the ratio -j has a value less 

than from 2 to 3, the lines of force spread out from the teeth 
and fill the upper end of the slots, forming a fairly uniform 
field in the air gap proper, and the actual density is very 
little increased on account of the slots removing a part of 
the armature surface. That this is practically correct is 
shown by the fact that pole-face eddy currents are negligible 



in such cases, as has already been stated 

greater than 3, the lines of force spread out after the manner 



If the ratio -j is 




PlO. 9 



of Fig. 9, but the density opposite the teeth is much greater 
than opposite the slot openings. 

43. For purposes of calculation, it has sometimes been 
assumed that the lines of force spread as in Fig. 10. In this 




FlO. 10 



case the approximate density may be taken as the quotient 
obtained by dividing the flux per pole by the average of the 



Digitized by VjOOQ IC 



is 



DYNAMOS AND DYNAMO DESIGIT g 14 



area of the pole face and of the tops of the teeth. This rule 
is simple and gives fair results when -j is about 4. For 

smaller values of -j it wil? give too high densities, and the 

ampere-turns for the air gap will be overestimated. A 
better rule is to take as the area of the air gap the sum of 
the areas of the teeth, assuming the width of each tooth to 
be increased by a strip A inches wide on both sides. Ordi- 
narily, no attention need be paid to the area of the teeth 
removed by ventilating ducts, or to differences in length 
between armature and field, as these will not greatly modify 
the values. 

44. In the design at hand, the teeth are . 95 inch wide 
at the top (see Fig. 4), and the air gap is \ inch, so the 
assumed equivalent width of a tooth is .95+(2xi) 
= 1.2 inches, and the length parallel to the shaft is 6} inches, 
or the area of the air gap per tooth is 1.2 X 6} = 8.1 square 
inches. Now, there are 7^ teeth opposite a pole, so the area 
to be taken for the air gap at each pole is 8.1 X 7} 
= 58.7 square inches. The above rule is to be used unless 

the ratio -j is less than, say, 2 or 2^, in which case the whole 

area of the armature beneath the pole is to be taken. 

The magnetic circuit as computed is made up as shown 
in Table II. 

TABIiB U 



Part 


Material 


Area of Path 
Square Inches 


Length of Path 
Inches 


Armature core . . 


Annealed punchings 


33.2 


10 


Teeth at roots.. 
Teeth at tops.... 


Annealed punchings 
Annealed punchings 


29.4 I Average 
36.6) 33 


2 X If = 3l 


Air gaps 


Air 


58.7 


2Xi = i 


Pole pieces 


Annealed punchings 


46.5 


2 X 9 = 18 


Yoke 


Steel casting 


30 


22) 





Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 



CX>MPUTATION' OF FIEIiD WINBINGS 

45, Magrnetization Curve. — To compute the field 
windings a magnetization curve, also called a saturation 
curve, is required. This is a curve showing the relation 
between the total flux per pole and the ampere-turns 
required to set up this flux. If the speed is fixed, the volts 
developed depend on the flux, and sometimes the curve is 
plotted between the volts developed and the ampere-turns. 
This last curve may be readily obtained experimentally 
from a completed dynamo, but it is more convenient for 
purposes of calculation to reduce the volts to flux per pole, 
for the armature may be run at other speeds than that of the 
test, and also other windings may be provided for other 
voltages, so that the volts would not be the same in all cases 
even though the flux per pole be kept at a certain desirable 
value. In other words, a change in the speed or armature 
winding would change a curve plotted between the volts and 
ampere-turns, but would not change a true saturation curve 
of a certain dynamo. These curves are very similar in 
shape to the magnetization curves already given for the 
various magnetic materials, only the latter are for a single 
material while the former are for a complete dynamo. 

To compute the magnetization curve for the foregoing 
dynamo, assume several values for the flux per pole and 
compute the ampere-turns required for each; plot the points 
on cross-section paper, and through them draw a smooth 
curve. The values assumed for the flux should be near that 
at which the machine is expected to run. 

46, The least number of points that will give satis- 
factory results is three, and if four or more are taken, the 
computations are probably correct if the points all lie on a 
smooth curve. More points should be taken if there is any 
doubt as to the shape of the curve within the working range. 
In the present design, assume successively for the armature 
flux per pole, 3,000,000, 3,500,000, and 4,000,000 lines. All 
the areas of the paths are known, and the densities, therefore, 



Digitized by VjOOQ IC 



30 



DYNAMOS AND DYNAMO DESIGN §14 



are readily obtained by division. The flux in the armature 
core is half of the total flux per pole; that in the magnet 
cores and pole pieces is 1.14 times the armature flux per pole 
on account of magnetic leakage; and the flux in the yoke is 
half of that in the magnet cores. The ampere-turns per inch 
of length of path are taken from the curves in Fig. 41, Part 2, 
except in the case of air, which is computed after the manner 
already explained. The several quantities used in obtaining 
the total ampere-turns are shown in Table III. 

47. These values are plotted in Fig. 11, and a smooth 
curve drawn through them. Now, the ampere-turns for the 




air gap for the several values of the flux varies directly as the 
flux, because the length and area of the path are constant. 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 



81 



TABIiE m 

For an Armature Flux of S, 000, 000 Lines Per Pole 



Part 


Flux 


Area 


Density 


IT 
Per Inch 


Length 


ToUl 
IT 


Armature core.. 
Teeth 


1,500,000 
3,000,000 

3,000,000 

3,420,000 
1,710,000 


33.2 
33.0 

58.7 

46.5 
30.0 


45.200 
90.900 

.51,100 

73,500 
57.000 


12 

44 
51.100 
3.192 

24 
20 


10 

18 
22i 


120 
154 


Air gaps 

Pole pieces. 

Yoke 


4,000 

432 
450 







Total ampere-turns for 3,000,000 lines 5.156 



For an Armature Flux of 8,600,000 Lines Per Pole 



Path 


Plus 


Area 


Density 


IT 
Per Inch 


Length 


Total 
IT 


Armature core. . 
Teeth 


1,750,000 
3,500,000 

3,500,000 

3,990,000 
1.995,000 


33.2 
33.0 

58.7 

46.5 
30.0 


52.700 
106,000 

59.600 

85,800 
66.500 


14 

95 
59.600 
3.192 

35 

27 


10 
3i 

i 

18 
22i 


140 
333 

4.667 

630 
608 


Air gaps 

Pole pieces 

Yoke 







Total ampere-turns for 3,500,000 lines 6,378 



For an Armature Flux of 4,000,000 Lines Per Pole 



Path 


Flux 


Area 


Density 


IT 

Per Inch 


Length 


ToUl 
IT 


Armature core.. 
Teeth 


2,000,000 
4,000,000 

4,000,000 

4,560,000 
2,280,000 


33.2 
33.0 

58.7 

46.5 
30.0 


60,200 
121,200 

68,200 

98,000 
76,000 


17 

240 

68,200 

3.192 

65 

40 


10 
3i 

i 

18 
22^ 


170 
840 

5.335 

1,170 
900 


Air gaps 

Pole pieces. 

Yoke 





Total ampere-turns for 4,000,000 lines 8,415 



Digitized by VjOOQ IC 



38 DYNAMOS AND DYNAMO DESIGN g U 

as is also the permeability of air; therefore, if any value of 
the air-gap ampere-turns is plotted, say, for 4,000,000 lines, 
5,335 ampere-turns, and a straight line drawn through this 
point and the origin (?, all other values of the air-gap turns 
for various fluxes may be read from this line. In Fig. 11, 
then, the ampere-turns to the left of the air-gap line are 
those required for the air gap, and those to the right, 
between the total curve and the air-gap line, are the ampere- 
turns required for the iron part of the magnetic circuit. 
Now, at low densities, iron conducts magnetism very readily, 
and very few ampere-turns are required, so the magnetiza- 
tion curve for small fluxes approaches the air-gap line and 
is tangent to it at zero flux. In drawing the magnetization 
curve, then, it must pass through the points plotted and be 
tangent to the air-gap line. 

48. When a saturation curve is obtained experimentally 
from the actual machine, it should be compared carefully 
with the calculated curves. Draw through the origin a tan- 
gent line for the air-gap line, and if this does not agree 
with that of the calculated curve, the method of estimating 
the area of the air gap should be modified in future calcula- 
tions. If the ampere-turns for the iron in the calculated 
curve are not correct, the magnetization curves of the 
materials may be at fault, and samples of the materials 
should be tested and new curves made. 

It is important that the magnetization curve of a dynamo 
be bent or curved, and this is accomplished by having some 
part or parts of the magnetic circuit magnetically saturated. 
If this is not done, a change in the ampere-turns will mean 
a corresponding change in the flux and voltage, and the 
latter becomes unstable ; while, if saturation takes place, a 
change in the ampere-turns will cause but a small change in 
the flux and voltage, and the machine is said to be stable. 
A good rule to follow to insure stability is to so proportion 
the magnetic circuit that, at full load, from 25 to 35 
per cent, of the ampere-turns is required for the iron. 
The full-load flux in the present design has been computed 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 33 

as 3,860,000 lines, which require by the curve some 
7,700 ampere-turns. For the air gap, 5,150 ampere-turns 
are required, leaving 2,550 ampere-turns for the iron, or 
about 33 per cent, of the total, 7,700 ampere-turns. This 
design should therefore be satisfactory as regards stability 
of voltage. 

EFFECTS OF ARMATURE REACTION 

49. Calculation of Cross Ampere-Turns. — Before 
taking up the design of the field windings, it is necessary to 
investigate the effects of armature reaction. The armature 
reaction depends on the load on the dynamo in kilowatts, 
and is the same for all three voltages, since the product of 
the total number of face conductors and the current in each 
is the same. The two effects of the armature reaction that 
are to be compensated for in the field windings are the back 
ampere- turns and the cross ampere-turns. These two have 
been shown to be equal,, numerically, to the ampere con- 
ductors within the double angle of lead of the brushes, and 
to those beneath the pole faces, respectively. To predeter- 
mine the angle of lead of the brushes necessary to secure 
sparkless commutation is difficult, if not impossible, so the 
value of the back ampere-turns is usually unknown until 
the generator is tested. However, they are much less 
important than the cross ampere-turns, because they are 
usually much smaller in value than the latter. The cross 
ampere-turns are computed by the formula 

ll> X Zi 
cross ampere-turns = — - — 

in which j^ = percentage of armature covered by poles; 
Z = total number of face conductors; 
i = current in each; 
p = number of poles. 

For the 500-volt armature, this is 

■ 75X464X100 ^ ^ ^^^ 

44—14 



Digitized by VjOOQIC 



34 DYNAMOS AND DYNAMO DESIGN §14 

60. Effect of Cross Ampere-Turns. — These ampere- 
turns tend to weaken one pole tip and strengthen the other 
to an equal degree. Referring to the table of ampere-turns 
for 4,000,000 lines flux per pole, which is about full-load 
flux, it, is seen that there are 5,335 ampere-turns required 
for the air gap and 840 ampere-turns for the teeth, or a 
total of 6,175 ampere-turns, to maintain the flux across the 
air-gap region if it is evenly distributed. If the ampere- 
turns required to set up this cross-flux in the armature core 
and pole pieces be neglected, the ampere-turns under the 
weak pole tips will be 6,175 — 5,800 = 375, while under the 
strong pole tips 6,175 + 5,800 = 11,975 ampere-turns will be 
displayed. Now, the density in the air gap will be propor- 
tionally weakened at the weak pole tips, but on account of 
the teeth having already reached saturation, the strong 
pole tips cannot be strengthened proportionally. Not only 
are the teeth saturated, but the pole horn was carefully 
designed to be strongly saturated: and, further, the air gap 
at the tips of the poles is | inch, while at the middle under 
the body of the pole it is but \ inch. Thus, in three ways, 
the cross ampere-turns are prevented from doubling the 
strength of field at the strong pole tips: (1) by saturation of 
the teeth, (2) by saturation of the pole horn, and (3) by 
increasing the air gap at the tips. Sometimes the punch- 
ings forming the pole pieces are made alternately long and 
short at the pole face, where magnetic saturation then takes 
place, and this has the same effect as saturation of the teeth. 
This method of constructing laminated poles is described 
later in connection with railway motors. 

51. Compensation for Cross Ampere-Turns, — Now, 
the effect of all this choking by saturation and otherwise at 
the^strong pole tip is to reduce the total flux, for the weak- 
ening effect of the cross ampere-turns is in no way pre- 
vented, while the strengthening effect is greatly interfered 
with. Thus, the cross ampere-turns tend to reduce the 
flux just the same as do the back ampere-turns, and in 
order to prevent the flux from being lessened, due to 



Digitized by VjOOQ IC 



§ 14 DYNAMOS AND DYNAMO DESIGN 35 

either effect, the machine should be provided with series- 
coils on the field magnets, that is to say, coils that are 
connected in series with the armature and through which 
all the current developed by the armature passes. Now, 
the armature reaction depends on the armature current, 
and since these series-turns are supplied with the same 
current, it is evident that by properly proportioning the 
number of turns in these series-coils, they may be made 
to satisfactorily compensate for both the back and cross 
ampere- turns. 

52. But these are not all the advantages to be derived 
from the saturation of the teeth, pole tips, etc. It was shown 
that, under the weak pole tips at full load, had there been 
no series ampere-turns added, only some 375 ampere-turns 
are impressed, against 6,175 ampere-turns at no load, so the 
strength of field would be practically zero. Now, a very 
weak field at this point is undesirable, making it necessary 
to rock the brushes far ahead to get the coils, under com- 
mutation, into the magnetic fringe of the pole, and thus 
assist in the reversal of the armature currents, as has been 
explained. The addition of ampere-turns on the field mag- 
nets by means of series-coils, which ampere-turns increase 
as the machine is loaded, cannot much affect the strong side 
of the pole that is already saturated, and the result is that 
the additional flux must find a path through the weak side 
of the pole. Consequently, the weaker side of the pole is 
greatly strengthened, and, instead of having one side very 
weak and the other very strong, as would have been the 
case without any magnetic saturation in the region near the 
air gap, a fairly good distribution of the flux over the pole 
face is obtained. While this is by no means a uniform dis- 
tribution, yet it is sufficiently near to it to be quite satis- 
factory. 

53. The saturation of the teeth, pole tips, and pole face, 
and the lengthening of the air gap under the tips, constitute 
the chief methods in use for overcoming armature reactions. 



Digitized by VjOOQIC 



36* DYNAMOS AND DYNAMO DESIGN §14 

They are incorporated in practically all modern generators, 
although some of these may have, in addition, some patented 
scheme for balancing the magnetic effects of the armature. 
The foregoing explanation of the action of the dynamo 
under load should be carefully gone over until thoroughly 
understood by the student and should be compared with the 
former explanations of armature reaction. It will be noted 
that, in designs wherein saturation of the iron at or near the 
air gaps is used for preventing the undesirable effects of 
the armature reaction, many more series ampere-turns are 
needed than in designs where no such saturation takes place. 
Therefore, it follows that the former type is quite unde- 
sirable for shunt-wound generators, because the voltage 
would fall off as the machine is loaded, on account of the 
choking of the flux at the strong pole tips, while the latter 
type, to which belonged most of the earlier designs of 
dynamos, operated quite satisfactorily as shunt-wound 
machines. 

54, Ampere-Tums to OfBSet Armature Reaction. 

From the foregoing discussion of armature reaction, it 
should be evident that the matter is too complex to com- 
pute the number of ampere-turns necessary to add as series 
ampere-turns to compensate for saturation. It is known, at 
least,- however, that they should be proportional to the 
armature ampere-turns, and from tests it has been found 
that, to compensate for both the cross and back ampere- 
turns, from 20 per cent, to 50 per cent, of the armature 
ampere-turns should be added to the series-coils of each 
magnetic circuit, the value depending on the degree to 
which the saturation is carried. In the design at hand, 
the armature ampere-turns are, from the equation, arma- 

,_ Zi 464X100 „ „^^ ^ , . 

ture //=—- = = 7,730 ampere-turns for the 

P 6 

500- volt windings. Supposing that 40 per cent, of these are 
allowed for armature reaction; then, about 3,100 ampere- 
turns will be required on the series-coils to compensate for 
the cross and back ampere-turns. 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 37 



CAIiCUI-ATTON OP FEBU) WINDING FOR 116-186 VOL.T8 

5^. For the purpose of illustrating the methods of cal- 
culating the field windings, let it be desired to wind a machine 
to give 115 volts at no load and 125 volts at full load, using 
the armature winding already determined for 125 volts. 
The variation of the voltage with the load is to be auto- 
matic, that is to say, the field-regulating rheostat is not to 
be adjusted. At no load, since there is no current flowing, 
there will be no resistance drop in the windings, so if 
115 volts is developed in the armature, there will be 
115 volts at the terminals. However, at full load there 
will be required about 8 volts to force the current through 
the windings; so, in order to obtain 125 volts at the termi- 
nals, about 133 volts must be developed in the armature 
windings. The flux per pole $ for no ioad is 

^ _ £ X 10* X 60 X /« „ 115 X 10" X 60 X 6 



ZxSxp 348X575X6 

= 3,450,000 lines approximately 

At full load the flux should be ??f ^^!^^ x 3,450,000 

115 volts * * 

= 3,990,000 lines per pole, because the voltage is directly 

proportional to the flux, the speed remaining constant. 

From the magnetization curve, Fig. 11, for 3,450,000 lines 

per pole, it is seen that 6,120 ampere-turns are required 

per magnetic circuit, and for 3,990,000 lines per pole, 

8,200 ampere-turns are required. Now at no load there is 

no current in the series-coils, so the shunt coils must be 

wound to develop the 6,120 ampere-turns. The regulating 

rheostat is adjusted to make the shunt current such that 

6,120 ampere-turns are developed at no load with 115 volts 

at the terminals of the shunt circuit. Now, when the 

machine is under full load, the terminal voltage is to be 

raised to 125 volts, so the shunt current will increase in the 

ratio ^H, and so will the ampere-turns developed by this 



Digitized by VjOOQ IC 



38 DYNAMOS AND DYNAMO DESIGN §14 

current. So the shunt ampere-turns at full load then are 
Ifl X 6,120 = 6,650, approximately. 

At full load, 8,200 ampere-turns are required per magnetic 
circuit without allowing for armature reaction, and 3,100 are 
needed to compensate for the latter, so that at full load the 
total ampere-turns on each magnetic circuit is 11,300. Of 
this, the shunt coils develop 6,650, so the series-coils must 
then develop the remainder, or 4,650. 

Referring to the diagram of the magnetic circuit. Fig. 5, 
it will be noticed that there are two sets of field coils on 
each circuit, so that each of the six shunt coils should 
develop 3,325 ampere-turns and each of the six series-coils 
2,325 ampere-turns at full load. 

66. Series- Winding:. — Taking up first the series- 
coils, the current for 100 kilowatts at 125 volts is ^VW^ 
= 800 amperes, so that each turn on the series-coils develops 
800 ampere-turns; hence, there will be required ^VV^» 
or 2.9 turns per coil. It is impossible to have a fraction of 

a turn of wire, so far as 
magnetic effects are 
concerned, for it is 
either a turn or not a 
turn. No matter what 
the direction of wind- 
ing, these coils must 
j^ alternate, since the 

p,Q ,2 poles alternate in sign, 

and the result of this 
is to divide the effect of one turn between two adjacent 
poles, so that it is usually stated that the turns per pole 
are 2^, 3^, or 4^, etc. Fig. 12 shows a winding of 1^ turns 
per pole. It will be noticed that between every two mag- 
net cores, there pass three conductors a, b^ and ^, or i, 2^ 
and ^, with the current in them in the same direction. 
This winding, therefore, has 3 turns per magnetic circuit, 
or 3 turns for two poles, but is usually said to have 1 J turns 
per pole. 



Digitized by VjOOQIC 



§14 DYNAMOS AND DYNAMO DESIGN 39 

57. In the design, 2.9 turns per coil are required, and 
there is the choice, therefore, between 2^ and 3 J turns on alter- 
nate poles, averaging 3 turns per pole, or of putting 3 J turns 
per pole and placing a shunt across the terminals of the series- 
coils to divert a part of the 800 amperes and thus reducing 
the series ampere-turns to any desired degree. Since the 
terminals of the series-coil must come out at opposite sides, 
it is not practicable to use 3 turns per coil ; the terminal must 
be brought around to the other side, thus making 3J turns. 

From similar designs, where the depth of the winding is 
Rot too great, about 1,200 circular mils per ampere has been 
found to give coils that will not heat over 40° C. above the 
temperature of the surrounding atmosphere; 800 amperes 
at 1,200 circular mils will require 960,000 circular mils, or 
960,000 X .7854 = 753,900 square mils, approximately, or 
about I square inch in section. In the diagram. Fig. 5, 
2 inches has been assumed for the depth of the space for the 
coil. Allowing ^ inch under and over coils for clearance and 
insulation leaves 1^ inches for the depth of the winding. 
We will use, then, a copper strip ^ in. x li in. bent on edge. 

68. For the winding room along the magnet core, it is 
necessary to count on the next greater number of turns, or 
4 turns. The conductor should be tapped with J-inch half- 
lapped tape, and, allowing .05 inch for insulation, this makes 
theconductor .55in. X 1.55 in. over all. Four turns along the 
core would require 2.2 inches or about 2 J inches for the con- 
ductor only; but there must also be left room for the con- 
necting leads from coil to coil. If ^ inch is allowed for each 
of these, the net length of the winding becomes 2} inches. 
Allowing ^^ inch for insulation under binding, and ^ inch 
for the diameter of the binding cord, adds ^ inch on a side, 
or I inch over all, so the total length of the series-coil along 
the core is 3J inches. 

69. Shunt Winding:. — The size of wire for the shunt 
winding is determined from the formula 

circular mils = 



Digitized by VjOOQIC 



40 



DYNAMOS AND DYNAMO DESIGN § U 



in which t T = ampere-turns per coil; 

L = mean length of a turn in inches; 
e = volts at the terminals. 



The ampere-turns required per coil are 3,325. The mean 
length of a turn may be estimated from Fig. 13. Allowing 

i inch all around the 



. _<cJle<w|_ l%*m_ 



I 

I 
I 
I 
I 
I 
I 
I 
I 

I 



^OuU*n4 QfMagnit Core 



a: 



.v.- 



41 



^J^ 



PlO. 18 



est turn is 39^ inches, 
shortest and the greatest is 



pole piece, which is 
7 in. X 7 in., makes 
the length of the short- 
est turn about 7^ X 4 
= 30 inches. The out- 
side turn will be longer 
than the inside by a cir- 
cumference of a circle 
whose radius is the 
depth of winding, say 
1^ inches. The circum- 
ference of a circle of 
1^ inches radius is 9.4 
inches or 9^ inches, so 
the length of the great- 
mean length between the 



== 34i inches. The 



The 

30 + 39i ^ 
2 

voltage at the terminals of each coil, since the shunt coils are 
connected in series, is one-sixth of the voltage across all the 
coils. In the shunt circuit, a regulating rheostat should be 
included, having a resistance to take up about 20 per cent, 
of the full-load voltage of the machine, leaving 80 per cent, 
for the coils. .80 X 125 = 100 volts for all coils, so that at 
the terminals of each coil there will be J|^ = 16.666 volts. 
Each shunt coil, then, must develop 3,325 ampere-turns 
with a terminal pressure of 16.666 volts and a mean length 
of turn of 34J inches. We have, then, circular mils 
3,325 X 34.75 



16.666 
wire table, it 



= 6,933 circular mils, 
will be seen that No. 



Referring to the 
11 B. & S. wire 



Digitized by VjOOQ IC 



gl4 DYNAMOS AND DYNAMO DESIGN 41 

has an area of 8,234 circular mils, and No. 12 B. & S. wire 
has an area of 6,530 circular mils. If no other sizes of 
wire than those of the B. & S. gauge are at hand, the coils 
could be wound with both sizes in such proportion as to 
bring the average circular mils, including the quantity of 
each size used, to the required value. If the coils in this 
case were wound entirely with No. 12 B. & S. wire, the 
voltage at the terminals of the coils would have to be some- 
what greater than 16f volts. How much greater is readily 

i Tx L 
determined by solving the equation, circular mils = , 

c 1. .u iTy^L 3,325X34.75 ^„ „ . 

for voltage; thus: ^= rp = - — :^^7^ = 17.7 volts; 

^ * cir.mils 6,530 ' 

so for the six coils in series, 106.2 volts would be required. 

This would leave 125 — 106.2 =18.8 volts for the rheostat. 

This is sufficient for regulating the voltage, and a winding 

of No. 12 B. & S. wire will therefore be used. 

60. The field wire has an area of 6,530 circular mils, so 
at 1,200 circular mils per ampere, the permissible current 
is \\l^ = 5.4 amperes, approximately, in the shunt circuit. 
With a current of 5.4 amperes the turns required per coil 

3 325 

will be -Vt- = 616, in order to give 3,325 ampere-turns. 
0.4 

The length of the winding space allowed on the sketch, 

Fig. 5, is 7 inches for both coils, and the series-coil has been 

found to require 8J inches of this, leaving 3| inches for the 

shunt coils. Allowing, as in the case of the series-coils, 

^ inch on a side for insulation and cord, or | inch for both 

sides, leaves 3 J inches for the net winding space. A single 

cotton covering on the wire will be sufficient for these coils, 

and from the wire table we find that No. 12 B. & S., S. C. C. 

has a diameter of .087 inch. A layer of this 3| inches wide 

should have 40 turns, or, say, 38 turns per layer in order to 

allow for clearance. Seventeen such layers would make a 

total of 046 turns per coil. The depth of Lhe winding is 

17 X .087 = 1.48 inches, which is very nearly the same 

as that for the series-coils. If the depth of winding is 



Digitized by VjOOQ IC 



42 DYNAMOS AND DYNAMO DESIGN §14 

considered too shallow, the winding space should be short- 
ened, making the diameter of the yoke less, while if it is too 
deep, the winding space must be lengthened. 

61. In the same manner, the field coils for any other 
voltage or any other compounding desired may be computed. 
The flux per pole at full load should always be about the 
same for a certain magnet frame, regardless of the com- 
pounding, so if it is desired to compound a generator for as 
great a rise as say 15 per cent, at full load, it is accomplished 
by lowering the flux at no load to a proper value rather than 
raising the flux at full load, and then selecting a suitable 
armature winding to give the voltage wanted at the required 
speed. 



TBTJE MECHANICAIi DESIGN 



«jtrMMARY OF dim:ensions 

62. The electrical design of the machine is complete, 
but the mechanical design still remains to be worked out. 
For convenience in referring to the quantities already 
determined that are involved in the mechanical design, the 
following data are given: 

Generator, 100 kilowatts; output 125, 250, 500 volts; 
575 revolutions per minute; belted type; 6 poles. 

Armature. — Diameter of armature 27 inches outside, 
11 inches inside; length of armature, 6J inches with one ^-inch 
air flue; number of slots, 58; size of slots, .51 in. X If in., 
with a ^-inch wedge in top for retaining winding in place. 

Armature Winding, 125 Volts. — Single parallel winding, 
one turn per coil, with four cross connecting rings; three 
coils per slot; 174 coils; coils span 10 slots; armature con- 
ductor, .10 in. X I in. 

Commuta7o>^ 125 Volts. — Nineteen inches diameter; face 
sufficient for six \" X If" brushes per set; 174 segments; 
coils connect to adjac?ijit segments. 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 4^ 

Armature Winding, 260 Volts. — Double series-windings, 
one turn per coil; four coils per slot; 232 coils; coils span 
10 slots; armature conductor, .075 in. x | in. 

Commutator, 250 Volts. — Nineteen inches diameter; face 
sufficient for four |" X 1|" carbons per set; 232 segments; 
coils span 78 segments. 

Armature Winding, 600 Volts. — Same as for 250 volts, 
but a single series-winding. 

Commutator, 600 Volts. — Nineteen inches diameter; face 
sufficient for three yV" X 1|" carbons per set; coils span 
77 segments. 

Poles, — Laminated iron; length, 7 inches, with |-inch end 
plates ; bore of poles, 27^^ inches ; angular span of poles, 45°. 

Magnet Cores. — Section, 7 inches square; winding room 
for field windings, 7 inches long by 2 inches deep. 

Yoke. — Cast steel; area of section, 30 square inches. 



DESIGN OP ARMATURE AND COMMUTATOR 

63. The armatures for the three voltages will differ only 
in the length of the face of the commutators, and for con- 
venience only that for 250 volts will be shown. The arma- 
ture will be designed with a spider upon which the commu- 
tator is supported, so that the whole may be made complete 
without the shaft, as is desirable for engine-type machines. 

64. Shaft and Spider. — The diameter of the shaft 
may be approximated from the formula already given, 



, 4/ w 



J. . / •/ watts 

diameter 



P. M. 
For 100 kilowatts, k is 1.2, so that 



=0^ 



100,000 , .^^ ,1 . u 

-3^—— = 4. 36, or, say, 4^ mches 



The hub of the spider should be heavy enough to with- 
stand the bursting strains when pressed on the shaft with an 



Digitized by VjOOQIC 




o 



Digitized by VjOOQ IC 



§ 14 DYNAMOS AND DYNAMO DESIGN 46 

hydraulic press. Supposing we use four arms about 2 inches 
thick for supporting the armature core, in one of which 
arms a key is inserted that fits into a nick in the punchings 
to prevent them from turning on the spider. Provision 
must be made for clamping the punchings together, and the 
method to be used will be plainly seen from the section of 
the armature shown in Fig. 14. The upper half of the end 
view shows the armature looking toward the commutator 
end ; the lower half shows the view looking toward the back 
of the armature. The front and rear end plates B and C 
are entirely separate from the spider A, The front end 
plate B strikes against a slight projection a on the spider 
arms, while the rear end plate C is keyed on by a small 
square key b^ bent into the form of a ring that fits into 
grooves cut in the spider arm and end plate. This key is 
made in two pieces, in order to insert it properly ; the joints 
are usually arranged at the spider arms, and are not seen in 
the completed armature. The punchings, of course, have a 
considerable spring, and when assembled on the core are 
clamped in a press, the key b inserted and the pressure 
released. The end plate C will then be pushed back and 
clamp the key tightly. 

The front end plate 5, as shown in Fig. 14, consists of a 
circular casting cored out and ribbed so as to obtain maximum 
stiffness for the least weight. 

65. Sometimes the front end plate is provided with 
flanges for supporting the winding, as shown for the rear 
plate C in Fig. 14, but in the present case the winding is 
made of heavy rectangular copper bars, and their stiffness, 
together with the support given by the necks h of the com- 
mutator segments, will be sufficient to properly support the 
winding. The rear end plate C is much the same design 
as B^ except that the flange c for supporting the winding is 
added. This flange is in turn supported by the eight arms d^ 
as shown in the rear end view. It will be noticed that in 
Fig. 14 there are provided very ample passages v up through 
the windings on the ends, and the arms d serve to fan air 



Digitized by VjOOQ IC 



46 DYNAMOS AND DYNAMO DESIGN g 14 

through these while running, and materially aid in cooling 
the armature. The distance the windings will project 
beyond the core is determined usually by making a careful 
drawing of the coils developed out into a plane instead of on 
a cylinder. The coils each Span ten slots, so the end con- 
nections from the core to the bend at the rear end should 
advance through five slots and another five in returning 
from the bend to the core. At the front end the advance 
on each commutator lead should be about five slots also, 
and the distance from the core to the commutator necks 
therefore should be about the same as that allowed for the 
end connections on the rear end. 

66. The ventilating flue n is formed by putting in a 
spacing disk, Fig. 15, consisting of an armature punching to 



XfVft 



U,H^ 



Fig. 15 



which sheet-iron strips have been riveted, as shown. These 
strips are ^ inch wide, the width of the ventilating flue, and 



Digitized by VjOOQIC 



§14 DYNAMOS AND DYNAMO DESIGN 47 

have two projections that are bent at right angles, one right 
and one left, thus forming feet by means of which the strips 
are held on edge. Through these feet small rivets pass for 
attaching the strip to the armature punching. 

67. The hub of the armature spider projects on the 
front end and is turned off to receive the commutator. 
This last consists of a spider or shell D and a clamping 
ring E arranged to hold the segments F between them. 
Both D and E are made of cast steel, as great strength is 
required to hold the segments tightly. The spider D is 
arranged with eight arms, leaving passages through which 
air can circulate between the commutator and the hub of 
the armature spider. The commutator is one of the most 
expensive parts of the dynamo, and is therefore made as 
small as practicable; this results in there being so little sur- 
face to radiate the heat developed that the temperature rise 
is usually greater here than on any other part of the 
machine. The ventilation of the commutator spider will 
materially decrease the temperature rise. The front end of 
the spider is turned off at e and the wedge ring is turned to 
fit this, as shown, the two parts being clamped together by 
the eight bolts G, These bolts are turned to a smaller diam- 
eter in the middle in order that there shall be considerable 
give to allow for the unequal expansion of the segments and 
the shell when heated. A suitable lock washer is held 
beneath the nuts of the clamping bolts to prevent them 
from loosening. 

68. In Fig. 16 is shown a detail of the commutator seg- 
ment. This consists of a hard-drawn copper bar F to which 
is riveted a neck h suitably arranged to receive the two ends 
of armature coils that terminate in the segment. The com- 
mutator should have sufficient face to accommodate four 
IJ-inch brushes. Allowing -^ inch between carbons, and 
^ inch for staggering the brushes, we have (4 X 1}) 
+ (4 X 7ff) = 7| inches as the actual working face of the 
commutator. Allowing | inch extra on the inside and :}inch 
outside, makes up the length 8J inches, as shown. Where 



Digitized by VjOOQ IC 



48 



DYNAMOS AND DYNAMO DESIGN 



§14 



the brushes do not butt against one another, that is, where 
there is a space between the carbons, the wear of the 



*j 



^- 



^ 




^ 






.»«»j 



-H — -^ 



^Copper. 
^30 Mica aegvMnU 



Pio. W 

carbons is liable to leave a ridge on the commutator between 
brushes, which may eventually necessitate turning the com- 
mutator in a lathe. To avoid this difficulty, the brushes of 
one brush-holder stud are placed out of line with those of 
the next stud in such a manner as to wear away any ridge 
that may be left by the first set. When brushes are so set 
they are said to be staggered. 

69. The segments should have sufficient depth to allow 
for |-inch wear, measured radially, so the point of the V cut 
should be, say, 1^^ inches below the face. Using the 
angles 5® and 30°, as shown, requires that the segments be 
about 2^ inches deep. The diameter at the surface is 
19 inches, and there are 232 bars, so the thickness of one 

. C A C ' U ^A u 3.1416 X 19 

segment of copper and one of mica should be — ;r 

2t52 

= .2572 inch. Taking the mica segment as. 030 inch thick, 

leaves .2272 inch for the thickness of the copper. The 

diameter of the segments at the bottom is 14 inches, so 

the thickness of one mica and one copper segment is 

-' = .1896 inch. Taking from this the thickness 

of the mica segment leaves for the copper .1596 inch, as 
shown. The copper is first drawn into bars of the proper 
section, then roughly cut to shape, after which they are 
milled to receive the neck A, which is both riveted and 
soldered into the ends of the segments. 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 49 

70. In assembling the commutator, the proper number 
of segments, with mica segments between, are held tightly 
together by clamps around the outside; then the V grooves 
are turned carefully to size. The micanite sleeve K^ Fig. 14, 
and the micanite cones H and /, are built up of thin leaves 
of mica pasted together with shellac or other varnish. These 
should be \ inch thick and made carefully to size. After the 
commutator is assembled, cord, shown at g, is wound over 
the mica cone on the outside end to protect it from injury. 
It is quite necessary that the insulating cones extend a con- 
siderable distance beyond the segments, for otherwise an 
accumulation of dust is liable to form a path for current 
over the mica, and the insulation between the windings and 
the frame of the machine will become impaired. 



pio. ir 

71. The commutator having been completed, it is forced 
on the armature spider, and is kept from turning by a key 
shown at w, Fig. 14. The armature is now ready to receive 

44—15 



Digitized by VjOOQ IC 



60 



DYNAMOS AND DYNAMO DESIGN 



§14 



the winding. For 125 volts, the winding is of the parallel 
type, with three coils per slot, a set of coils for one slot being 
shown in Fig. 17. In the diagram, the armature core is, for 
simplicity, shown as straight instead of circular. The coils 
span 10 slots, that is, they lie in slots 1 and 11, the lower 
side A being in slot i, the upper side B in slot 11. Beneath B 
is the bottom side of some other coil C, and in the completed 
armature, each slot will have a set of conductors in the top 
and a set in the bottom. At the end A the three conductors 
spread out and are connected to the segments i, 2, and 3, 
while the ends B connect to the segments 2, S, and ^, so that 
coil 1 begins at segment 1 and ends at segment 2, It will 
thus be seen that each coil terminates in adjacent segments, 
which is the essential feature of the single parallel winding. 




Fig. 18 



72. For 250 and 500 volts, the armature coils are of the 
series-type, and have the appearance indicated in the dia- 
gram, Fig. 18; in this winding there are four coils per slot 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 51 

taped together, and, as before, they span 10 slots. The 
necks of the commutator segments are omitted in order to 
avoid confusion, but it will be understood that the ends A 
go into the lower part of the necks and the ends B in the 
upper part of the necks, in the same manner as before. In 
the diagram, which is for the 250-volt winding, coil 1 con- 
nects from segment 1 to segment 7P, spanning 78 segments. 
From segment 79^ another coil connects, spanning 78 more 
segments, and terminating in segment 157; another coil 
connects this to segment 157 + 78 = 235, or, since there are 
but 232 segments in the commutator, segment 235 is seg- 
ment S. Thus, a series of ^, or three coils, / being the 

number of poles, extends around the commutator, termina- 
ting in segments 1 and S or in segments adjacent but one, 
which is the distinguishing feature of the double series- 
winding. 

73. For 500 volts, the coils are very similar to Fig. 18, 
but should span only 77 segments, so that the ends at B 
should be marked to segments 78, 79^ 80^ and 81, instead of 
7P, 80, 81, and 82, as shown. If each coil spans 77 seg- 
ments, a series of three would span 231, or starting at seg- 
ment 1, would terminate in segment 282, which is adjacent 
to segment i. This, then, would be a single series-winding. 



CONSTRUCTION OP FIELD FRAME AND FIELD COILS 

74. The field frame shown in Fig. 19 is divided horizon- 
tally into two parts for convenience in assembling or repairing 
the generator. It will be noticed that the area of the sec- 
tion of the yoke is greater than the calculated value required. 
It is frequently found necessary to considerably increase the 
area of a cast-steel yoke over that required for the magnetic 
flux, on account of the section being to6 small to be suffi- 
ciently stiff to withstand the powerful pull of the magnets. 
If the frame springs any, the length of the air gap at the 
various poles becomes unequal, and the strength of these 



Digitized by LjOOQ IC 



52 



DYNAMOS AND DYNAMO DESIGN §14 




Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 63 

poles is then affected so as to tend to distort the yoke still 
further out of shape. In the present case, the density 
assumed for the yoke calls for a cross-section of 30 square 
inches. This would make a rather light frame, and the 
section shown in Fig. 19 has an area of about 43 square 
inches, which would be better for a machine of this size, 
simply on account of the mechanical considerations just 
mentioned. This increase in the yoke section will not 
affect the field windings appreciably. The cross-section has 
been increased a little over 40 per cent. The yoke was 
originally calculated for a density of 75,000, so that the- 
density with the increased cross-section would be about 
H X 75,000 = 52,326. The ampere-turns required for the 
yoke are only about 10 per cent, of the total ampere-turns, 
and a reduction of the yoke density from 75,000 to 52,326 
would correspond to a reduction of not more than 3 per 
cent, in the total ampere-turns. A change of one size in 
the shunt wire means a change of over 20 per cent, in the 
total ampere-turns, so that it is easily seen that the increase 
in the yoke section does not call for a recalculation of the 
magnetic circuit. The yoke density is comparatively low, 
so that an increase of cross-section has comparatively little 
influence on the field windings. 

76. On the inside of the yoke are shown six pads a that 
are to be machined off to receive the pole pieces that are 
to be bolted to the yoke by two tap bolts, as shown. Two 
poles with their magnetizing coils are shown in place. On 
the front side of the yoke are shown four pads b that are to 
be machined off to receive four brackets for supporting the 
rocker-arm, details of which will be shown later. A pad c on 
the upper half is for attaching an adjusting screw by means 
of which the position of the brushes can be changed. In 
order to allow a little more room for the field coils, the bore 
of the pole-piece seats a has been increased to 45 inches, the 
inside diameter of the yoke to 45^ inches, and the outside 
diameter to 58^ inches. These dimensions are \ inch greater 
than those taken in Fig. 5, but this slight change will not 
affect the magnetic calculations perceptibly. 



Digitized by VjOOQ IC 



54 



DYNAMOS AND DYNAMO DESIGN §14 



In each of the two feet are shown two holes e for bolts for 
attaching the yoke to the bedplate. In addition to these, 
there is a single hole / in each, for a taper pin. In assem- 
bling the machine, after the correct position of the yoke on 
the bedplate has been found, the bolts are tightened and 
the taper pin adjusted carefully and accurately into place. 
If the machine is taken apart, when it is reassembled the 
taper pins are inserted and driven into place first, and then 
the bolts put in and screwed up. . By this means the yoke 
will be made to take its correct position on the bedplate 
without any other adjusting. 

76. The details of the pole pieces are shown in Fig. 20* 
They consist of punchings ^ inch thick, riveted between 




FlO. 20 

two |-inch wrought-iron plates by five rivets. One of these 
is \\ inches in diameter, and this is drilled and tapped for 
two |-inch bolts that serve to hold the pole pieces to the 
yoke. The corners of the end plates should be carefully 
rounded, so that the insulation of the field coils will not be 
injured by the sharp edges. 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 55 

77. In Fig. 21 the details of the field coils are shown. 
Both shunt coils and series-coils are wound on forms to leave 
7^ inches clear inside each way, the corners being bent to a 
radius of | inch, as called for in Fig. 13. The field windings 
for 125 volts are shown, as these only have been computed, 
but the coils will remain practically the same size for what- 
ever voltage is used. The shunt coil is wound with No. 12 
B. & S. single cotton-covered wire, with 38 turns in a layer, 




Dord 

T 



-X9 ihUknem 
Sheet Copper 



Pio. 21 



and 17 layers. After the coil is wound, it is necessary to 
provide it with suitable terminals. If the end of the wire is 
left projecting for the terminal, it is liable to be broken off, 
and if the inner end of a coil that is wound on a spool is 
broken, the whole coil may have to be rewound. To avoid 
this, the terminals of the coils are often made of short pieces 
of flexible wire made of many strands of fine copper wire, 
which is not so easily broken. In the present case, a ter- 
minal with binding screws for attaching the leads from the 



Digitized by VjOOQ IC 



56 DYNAMOS AND DYNAMO DESIGN §14 

coils has been used. Such a terminal itself must be strong 
and firmly attached to the coil, as the wire is quite heavy, 
and even with the best handling, a projecting terminal is 
liable to receive a severe knock. The details of the coil 
terminal will be understood from Fig. 22. It consists of a 
small brass casting a^ with a broad flat foot that sets against 



PlO. 2S 

the coil. To the bottom of this casting is soldered a strip 
of copper ^, to which in turn the end of the wire is soldered. 
Beneath the strip and casting is a thin sandwich c of 
press paper and oiled muslin to protect the winding and 
insulate it. 

78, In putting on the terminals of the coils, great care 
must be taken to properly insujate them, as a failure of this 
insulation will leave the coil wholly or partially short-cir- 
cuited. The voltage impressed between the terminals of a 
coil during its normal operation is not the only voltage it is 
required to withstand, for the coil surrounds a magnetic 
flux and any change in the amount of flux threading a coil 
will induce in it an E. M. F. due to its self-induction. 
Suppose the field circuit should be suddenly broken while 
there was current flowing in the coil. The flux in dying 
away would cause the voltage at the coil terminals to 
become several times as great as its normal value. This 
potential will be greatest at the terminals of the coil, and, 
referring to Fig. 22, it will be noticed that the terminal 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 67 

shown connects to the under side or inner end of the coil, 
while the terminal itself rests on the wires of the outer 
layer. Between the two sides of the insulation c, then, there 
may exist a considerable voltage, and it is important, there- 
fore, that this insulation be good and reliable. 

79, After the terminals are in place they should be 
bound firmly to the coil with small flax cord, and then the 
complete coil should be insulated with a wrapping of oiled 
muslin, and over this a wrapping of cotton tape. The coil 
is next bound together at eight points with |-inch cord 
wrapped over a strip of sandwich of oiled muslin and press 
paper. This cord forms the chief protection for the coil, as 
it will be seen that the cord alone comes in contact with 
adjacent parts. 

Formerly, it was customary to wind field coils on spools, 
but this practice is not greatly used today, as the spools are 
expensive, and the extra protection they afford is considered 
unnecessary, since the coils should not in their normal use 
be subjected to such handling as to require better protection 
than is given by the method just described. 

80, The terminals of the series-coils consist, as shown in 
Fig. 21, of 12 thicknesses of sheet-copper strip 3 inches 
wide, and .020 inch, or No. 24 B. & S., thick. The total 
area of this terminal is about the equivalent of the series- 
coil copper, which is J in. x 1 J in. In attaching these leads 
to the coil proper, the copper strips should be both riveted 
and soldered carefully to insure a good electrical connection. 
Countersunk rivets should be used, and it will be necessary 
to place over the strips a piece of hard sheet brass ^ in. 
thick X 3 in. X 1^ in., in order to have some solid metal to 
hold the head of the rivet. 

81, The method of connecting two series-coils that are 
provided with this kind of terminal will be readily under- 
stood from Fig. 23. The leads from one coil are interleaved 
with those from the other, and the whole clamped together, 
between two small brass plates, by two bolts. This joint is 



Digitized by VjOOQ IC 



58 DYNAMOS AND DYNAMO DESIGN §14 

excellent from an electrical point of view, for the surfaces in 



mn 




^ 



Pig. 28 

contact have so great an area that the current density in the 
contact surface is small, as is also the resistance of the joint. 



BRUSH HOLDERS AND ROCKER 

8!3, In Fig. 24 the details of the brush holder are shown. 
Each of these consists of a light cast-brass pocket ^, in 




CO , 




PIO. 24 



which the carbon brush is capable of sliding readily. This 
pocket is supported by two flat springs b^ which in turn are 



Digitized by VjOOQIC 



gU DYNAMOS AND DYNAMO DESIGN 69 




Digitized by VjOOQIC 



60 DYNAMOS AND DYNAMO DESIGN § U 

attached to another casting c that is drilled for mounting on 
the brush-holder stud. To the casting c a flat flexible 
spring d is attached, one end of which presses against the 
carbon brush and keeps it in contact with the commutator. 
An adjusting screw ^is provided for regulating the pressure 
on the brushes. It is very undesirable to have those parts 
of the brush holder carry current that are also used as 
springs, and it is especially undesirable to have to rely on 
the sliding contact between the brush and the metal pocket 
for carrying current; so, to each carbon there is attached a 
short flexible copper conductor, termed a pigtail, which in 
turn is attached to ^/ at a point where there is ample cross- 
section for carrying the current, and therefore little likeli- 
hood of sufficient heating to draw the temper of the spring. 

83. For the 250-volt generator there are four brush 
holders per stud. In Fig. 25 is shown the rocker-arm and 
the supports for the brush-holder studs. The brush-holder 
studs consist of a rod of brass | inch in diameter, which is 
supported at either end by an iron casting A. These cast- 
ings are in turn fastened to the rocker-arm or ring by four 
screws, but are insulated from the ring by vulcabeston 
washers and bushings, as shown at d. This construction 
affords a firm support for the brush holders, and insulates 
them thoroughly. To the rocker ring are also attached six 
maple cross-pieces c, which support at their ends two copper 
bus-rings D and £, one positive and one negative, to which 
the brush-holder studs are connected, as shown at /, by 
short flexible cables. To these bus-rings are also attached 
the armature leads // and ^, which serve to connect the arma- 
ture with the outside circuit. On the outside edge of the 
ring will be noticed four short lips, or ribs, /«, that are 
machined to fit into groove in four arms, shown in Fig. 28, 
which serve to support the rocker-arm and attach it to the 
frame. At w, Fig. 25, is shown a hole drilled to receive a 
stud into which the adjusting screw, Fig. 28, is fastened, so 
that, by turning the hand wheel, the position of all brushes 
on the commutator can be adjusted at will. In many of the 



Digitized by VjOOQ IC 




o 



Digitized by VjOOQ IC 



62 DYNAMOS AND DYNAMO DESIGN §14 

earlier generators, the rocker-arm was attached to the bear- 
ing rather than to the frame of the machine. This con- 
struction has been abandoned on engine-type generators, 
for the reason that it is customary for the engine builder to 
supply the outboard bearing and pedestal, and the rocker- 
arm as formerly constructed would have to be specially 
fitted to the pedestal. The use of a ring-type rocker 
attached to the magnet frame is not open to this objection, 
the engine-type machine being complete in itself. The use 
of this construction for engine-type apparatus has led to its 
adoption on many of the belted types, as it is an advantage 
commercially to use the same parts for both types thus 
avoiding the multiplicity of parts. 



BEDPIiATE ANB BEARINGS 

84. In Fig. 26 the details of the bedplate and bearings 
are shown. The bedplate is cast in one piece, with the two 
pedestals, forming a construction that is of neat appearance 
and that avoids the machining of joints between the pedes- 
tals and bed. The casting, however, is complicated, and 
some manufacturers prefer making the bedplate separate, 
bolting the pedestals into place. In machines of larger size 
than that shown, this latter construction becomes essential, 
as the difficulty of machining a large casting, with pedestals 
cast on, outweighs the advantages of the single casting. 

86, The bedplate as shown consists of a thin shell of 
cast iron ribbed at points to strengthen it. Such a bedplate 
has an appearance of solidity, and has ample strength and 
stiffness without requiring great weight. The pedestals are 
provided with caps that are bolted into place, and these 
hold the bearing, as shown in the section. The bearing 
itself is of the spherical type, so that it is capable of slight 
adjustment in any direction in order that the two bearings 
may adjust themselves properly in line. The bearing con- 
sists of cast iron, and is lined with babbitt. Oil rings tf, tf, 
that run on the shaft are provided, and grooves are cut in 



Digitized by VjOOQ IC 



Digitized by VjOOQIC 




Fig. 87 



Digitized by VjOOQIC 




Digitized by VjOOQIC 



Digitized by VjOOQIC 



§14 DYNAMOS AND DYNAMO DESIGN 63 

the bearing shell to allow the rings to hang on the shaft and 
dip into lubricating oil retained in the reservoir provided in 
the pedestal. As the shaft revolves, the oil is carried up 
into the bearing by means of the rings. The oil then flows 
along the shaft in suitably cut grooves in the bearing, 
escaping back into the reservoir through the holes shown 
at b. During the operation of the machine there is a con- 
tinuous flow of oil through the bearing, which not only keeps 
the shaft well lubricated, but tends to wash away any par- 
ticles of metal that may be worn from the rubbing surfaces. 
In order to prevent the bearing from turning in the pedestal, 
a slot is cut in the former and a setscrew c that projects into 
the slot is provided in the cap. Two holes are provided in 
the cap, directly over the two oil rings, for the purpose of 
inspecting the rings to see if they are revolving freely, and 
also for introducing oil. These holes are supplied with two 
small covers d^ as shown. The height of the oil in the oil 
reservoir should be such that the rings dip well into it, but 
should not be so high as to run out of the bearing beside the 
shaft. An oil gauge is attached at e for indicating the 
height of the oil in the reservoir; the gauge is shown in 
place in Fig. 28. 

86, It is important that the oil from the bearings should 
not follow the shaft and run out on the armature. The oil 
itself is not so injurious, although it tends to destroy the 
insulating varnishes, but it causes an undesirable accumula- 
tion of dust that may greatly impair the insulation. It is 
to prevent the oil from reaching the end of the bearing that 
the oil grooves and the end holes b are provided. An oil 
groove is cut on the shaft, as shown at /, Fig. 26, so that 
what little oil passes out from the end of the bearing is. 
thrown off from the sharp edge of /"by the centrifugal force 
and falls within the casing. 

87, Figs. 27 and 28 show the general assembly of the 
machine. The former shows a half section, and the rela- 
tive positions of the various parts can be readily seen. The 
shaft is provided with two keys, one in the hub of the 



Digitized by VjOOQ IC 



64 DYNAMOS AND DYNAMO DESIGN §14 

armature spider and the other in the hub of a pulley. The 
diameter of the pulley is 28 inches, and that of the armature 
is 27 inches, so the belt speed will be just a little higher than 
the peripheral speed of the armature, which was designed 
for 4,000 feet per minute. This speed is about right, although 
it could be as high as 5,000 feet, with the advantage that the 
belt pull would be proportionally decreased. A low belt 
speed is objectionable on account of the increased pull neces- 
sary to transmit the power required. It will be noticed that 
the rocker-arm, with the brush holders, are shown in place, 
and the construction of these parts will be fully understood 
from Figs. 27 and 28. 

Fig. 28 gives, perhaps, the best idea of the complete 
machine. The oil gauge is in place on the pedestal. This 
consists of a glass tube protected by a brass sheath, through 
which a slot is cut to permit the height of the oil being seen. 
A small cock is provided for emptying the reservoir of oil. 
The construction of the adjusting screw for the locker-arm 
will be understood from the figure. All other parts shown 
have been previously detailed and described. 



CONITECTIONS 

88, In Fig. 29 is shown a diagram of the connections of 
the machine. The current enters the machine from the 
negative lead, as indicated by arrows, and passes from the 
terminal board A directly to the negative bus-ring B on 
the rocker-arm. From this bus-ring it divides between the 
three brushes C connected thereto, going into the armature 
and out again at the brushes D, which are connected to the 
positive bus-ring £, From £ the current passes through 
another flexible cable to the positive terminal board i% in 
which the leads to the series-field winding terminate; the 
current then passes through the series-coils and out on the 
positive lead. Connecting to the terminal board F will be 
noticed another outside connection, termed the equalizer 
lead. This connection is required where several compound- 
wound generators are operated in parallel, in order to cause 



Digitized by VjOOQ IC 



Digitized by VjOOQIC 



Pig 



Digitized by VjOOQIC 



Fic. 16 



Digitized by VjOOQIC 



Digitized by VjOOQIC 



§14 



DYNAMOS AND DYNAMO DESIGN 



66 



the line current to properly divide between them. The use 
of the equalizer connection is fully described in the section 




PIO. 89 

relating to the operation of dynamos in parallel. When two 
or more machines are run in parallel, the equalizer wire con- 
nects together those brush terminals to which the series- 
coils are attached. 

89. In Fig. 29 one end of the shunt-field circuit is con- 
nected to the armature lead on terminal board A and the 

44—16 



Digitized by VjOOQ IC 



66 



DYNAMOS AND DYNAMO DESIGN 



§U 



other end to a small terminal (7, from which a lead wire runs 
to the field-regulating rheostat H that is connected in series 
with the shunt coils, the circuit being completed by a con- 
necting wire to the positive line; this last connection is 
usually made on the switchboard. The direction of the 
currents in the coils can now be followed, and is as indicated 
by the arrows. It will be seen that the magnetizing effect 
of both the shunt coils and the series-coil on each pole is in 
the same direction, and inasmuch as the poles alternate N 
and 5, it is necessary that the current in the coils on 
adjacent pole pieces should flow in opposite directions. 





PlO. 80 



90. The details of the positive and negative terminal 
boards are shown in Figs. 30 and 31. The flexible cables are 



Digitized by VjOOQ IC 



§14 



DYNAMOS AND DYNAMO DESIGN 



67 




soldered into cast-brass terminals that are provided with flat 
parts of ample area, so that when bolted or clamped together, 
or to another con- 
necting piece, the 
current ^density in 
the area in contact 
shall not exceed 
from 80 to 125 am- 
peres per square 
inch. These ter- 
minals are clamped 
as shown by small 
bolts a that pass 
through the slate 
block and screw into p»g. 81 

small brass blocks *, which serve to support the terminals 
away from the slate. On the positive terminal board, these 
blocks are milled out at //, to take the flat copper strip that 
connects to the series-coils. The slate blocks themselves are 
each attached to the frame of the magnet by four tap bolts r, 
and it is necessary to support the terminal blocks away from 
the frame by ferrules d^ as shown, in order to insure that the 
screws and bolts, that attach parts carrying current and that 
pass through the slate, do not come into contact with the 
frame and thus ground or short-circuit the generator. It will 
be noticed that the heads of the bolts c that attach the termi- 
nal boards to the frame are sunk into the slate ; this is done 
in order to prevent accidental connection between a terminal 
and these bolts with a wrench or other tool. 



EFFICIENCY 



1S5-VOI.T GENERATOR 

91. The separate losses of a dynamo may be measured 
or computed with considerable accuracy, and it is therefore 
permissible to compute the efficiency from these losses. As 
an example of the calculations, the efficiency of the fore- 
going design will be given. 



Digitized by VjOOQIC 



68 DYNAMOS AND DYNAMO DESIGN §U 

The losses in a dynamo consist of two kinds, mechanical 
and electrical. The mechanical losses are those due to 
bearing friction and windage^ commutator brush friction^ and 
the iron losses. The electrical losses consist of the armature 
PR loss, the commutator PR loss, the series-field PR loss, 
and the shunt field PR loss, including that in the rheostat. 
Consider the 125-volt machine, and let us determine the 
above losses in the order they are given. 

92. Bearing: friction and wlndagre may, on the size 
given, be taken as about 1 per cent, of the output, or 
1,000 watts. 

93. Commutator Brush Friction. — There are required 
6 X 6, or 3G carbons \ in. X IJ in., and these rub on a com- 
mutator 19 inches in diameter, which runs at 575 revolutions 
per minute. At this speed the friction loss may be taken as 
50 watts per brush, or for 36 carbons, a total of 1,800 watts. 

94. The armature I' R loss can be computed from 
the current and the resistance. The current for 100 kilo- 
watts at 125 volts is 800 amperes. The /' R loss in the slots 
has already been computed, but for the present it is neces- 
sary to include the end connections as well as the face con- 
ductors. The area of the face conductor, .1 in. x | in., has 
already been determined as 76,900 circular mils, and the 
length of a coil may be taken as 52 inches. The resistance 

.... . r, length in inches 

of a coil when warm is A = ? : -, rp, or 

area in circular mils 

R = jy^-57^ = .000676 ohm. The resistance of the armature 

r C 
is R =z -^, where r- is the resistance of a coil ; C, the num- 
7n' ' 

her of coils; and ;;/, the number of paths or circuits through 

the winding, r^ in this case is .000676 ohm, C is 174, and 

;// is 6, so that R = '^^^^''^f.^ ^'^^ = .0033 ohm. The arma- 

ture P Rloss, then, is (800)' X .0033 = 2,112 watts. 

The resistance of the contact surfaces on the commutator 
is such that the drop is about 3 volts at the ordinary current 



Digitized by VjOOQIC 



§14 DYNAMOS AND DYNAMO DESIGN 69 

densities used in carbon brushes. At 800 amperes, this 
would mean a commutator I* R loss of 2,400 watts. 

95. lioss in Series-Field. — The series-field conductor 

is a copper bar ^ in. X 1^ in., and has been shown to 

have an area of 960,000 circular mils. There are 3^ turns 

per coil, and 6 coils, and each turn is 34f inches long, 

,, . , . ,, .... 34.75 X 3.5 X 6 

so the resistance of the series-coils is 

ybu, uuu 

= .00076 ohm. Allowing for the connections, it will be safe 

to call the resistance of the series-circuit .0009 ohm, so the 

loss in the circuit is (800)' X .0009 = 576 watts. 

96. liOSs in Sliunt Colls. — The current required in the 
shunt coils was estimated at 5-r4 amperes, so the loss in 
this circuit, including that in the regulating rheostat, is 
5.4 amperes X 125 volts, or 675 watts. 

97. Summary. — The iron losses in the armature core 
have already been estimated as 2,040 watts. 

The mechanical losses are as follows: 

Bearing friction and windage 1,000 watts 

Commutator brush friction 1,800 watts 

Iron losses in armature core 2,040 watts 

Total 4,840 watts 

The electrical losses are: 

Armature PR loss 2,112 watts 

Commutator PR loss 2,400 watts 

Series-field PR loss 576 watts 

Shunt-field PR loss 675 watts 

Total 5,763 watts 

If the generator delivers 100,000 watts at its terminals, 
and there are 5,763 watts electrically lost, then there must 
have been developed 105,763 watts in the winding, and as 
shown in Part 1, 

-, W 100,000 ^, ^ 
^•=T*^= 105:763 = ^^-^P^^^^^^ 



Digitized by VjOOQIC 



70 DYNAMOS AND DYNAMO DESIGN § 14 

which is the electrical efficiency, where IV is the output and 
Wf is the internally developed watts. 

98. The efficiency of conversion [7^ has been defined as 
the ratio of the internal electrical watts Wi to the total watts 
supplied to the belt IV^. That is, 

There are 4,840 watts lost mechanically and 6,763 elec- 
trically, so the total losses are 10,603 watts, or IV, 
= 110,603 watts. Hence, 

„ W, 106,763 -_. 

^• = 7n = no;603 = ^^-^P"'" """'•' 

which is the efficiency of conversion. 

The real efficiency, or the commercial efficiency, 17 has 
been defined as the product 

rr rr ^ rr ^ 100,000 

^= C/cXC/. = -^ = 3^^^= 90.4percent. 



«50-VOIiT AiTD 500-VOI.T GENERATORS 

99. The foregoing figures are for the 125- volt generator 
only, and it will be found that the 250-volt machine will 
have a higher efficiency, and the 500-volt machine an effi- 
ciency still higher. Let us consider how these machines 
differ. They all have the same bearings, bedplate, and 
armature core, and the bearing friction and windage will be 
the same on all. The magnet frame is the same, and the 
field windings will require about the same number of watts 
to set up the flux regardless of the voltage. The armature, 
while provided with different windings, has about the same 
number of circular mils per ampere and the same total 
ampere conductors, so that the total weight of copper 
and also the total f^R loss will remain about the same for 
all. The commutators, however, are different for the three 
voltages, and the change in efficiency is due wholly to the 
difference in the losses in the three commutators. 



Digitized by VjOOQ IC 



g U DYNAMOS AND DYNAMO DESIGN 71 

100. Considering the commutator losses as computed 
for the 125-volt generator, the friction of 36 carbons was 
estimated as 1,800 watts, and the PR loss for 800 amperes 
as 2,400 watts, or the total commutator loss is 4,200 watts, 
which is 40 per cent, of the total losses. For 250 volts, the 
current is 400 amperes, and there are but 24 brushes on the 
commutator, so the PR loss in this case is 1,200 watts and 
the friction loss 1,200 watts, making a total of 2,400 watts 
for the total commutator loss. This is 1,800 watts less than 
for the 125-volt machine, and computing the efficiency with 
1,800 watts less loss than for the 125-volt machine, we 
obtain a commercial efficiency of 92 per cent. 

101« In the same manner, for 500 volts there are but 
18 brushes, so the friction loss is about 900 watts, while the 
current is but 200 amperes, so the PR loss is 600 watts, 
making a total of 1,500 watts. This is 2,700 watts less than 
the loss for 125 volts. In this case, the commercial effi- 
ciency is 92.7 per cent. 

102. It is often desired to know the efficiency of a gen- 
erator at loads other than full load. A machine might have a 
very high efficiency at full load and yet at lighter loads have 
such low efficiencies as to be less economical than a machine 
with a high efficiency at lighter loads although having a lower 
efficiency at full load. The average generator in ordinary 
use runs at about one-half to three-fourths load most of the 
time, with an occasional greater load, so that the efficiency 
at full load is often not so important as that at lighter loads. 
As an example, we will compute the efficiency of the 125-volt 
generator at ;J^, i, and f load. For this purpose, it is neces- 
sary to separate the losses that change with the load from 
those that do not. None of the mechanical losses change 
with the load to any extent, while all the electrical losses, 
except that of the shunt field, do change. 

The constant losses are : 

Mechanical losses 4,840 watts 

Shunt-field loss 675 watts 

Total constant loss 5,515 watts 



Digitized by VjOOQIC 



^'2 DYNAMOS AND DYNAMO DESIGN § U 

The variable losses are : 

Armature loss 2,112 watts 

Commutator loss 2,400 watts 

Series-field loss 576 watts 

Total variable loss 5,088 watts 

These variable losses depend on the square of the current, 
so that for one-quarter load they will be ^ X }, or ^ of the 
full load value, and at one-half load they will be ^ of full 
load, and at three-quarters load ^ of full load. 

Thus: 

\ Load ^ Load | Load 

Constant losses 5,516 5,515 5,515 

Variable losses 318 1,272 2,862 

Total losses .... 5,833 6,787 8,377 

Output 25,000 50,000 75,000 

Input 30,833 56,787 83,377 

Efficiency (percent.) 81.08 88 89.95 

It is seen that the efficiency at one-quarter load is quite 
high, and is practically constant from one-half load to full 
load, being between 88 and 90 per cent. 



TESTING 

103* After completion, dynamos should be subjected to 
rigid tests to determine whether they are perfect and 
whether they are suitable for the service for which they are 
intended. The character of the tests should depend on what 
the machine is required to do. In order to insure getting 
machines of sufficient capacity, customers usually require 
that the machine shall be tested in a manner that they pre- 
scribe, for a certain length of time, and that its performance 
under test shall come up to certain desirable standards. The 
chief requirements are in regard to the rise in temperature, 
sparking, and adjustment of the brushes, efficiency, capacity 
to stand overloads, compounding, and the insulation of the 



Digitized by VjOOQ IC 



§ 14 DYNAMOS AND DYNAMO DESIGN 73 

windings. The usual requirements in each case will be 
taken up separately. 

* 104. Rise in Temperature. — It is usually required for 
constant-potential generators, intended for electric lighting 
or power, that the rise in temperature after a continuous 
run at full load shall not exceed 40*^ C, measured by ther- 
mometer on any part except the commutator, which is per- 
mitted to rise from 50° to 55° C. The temperatures are 
taken by thermometers placed upon the outside of the part 
of which the temperature is desired, and the bulb of the 
thermometer should be protected from drafts by a small 
piece of waste or other non-conductor of heat. It is further 
specified usually that the room temperature shall be either 
at 25° C, or 77° F., or referred to that temperature. 

105. In the case of the windings, the change in resistance 
of circuits of copper conductors may be used to measure the 
temperature rise, for it is found that an increase in tempera- 
ture of copper will increase its resistance 1 per cent, for 
every 2^° C. above 25°. If the rise in temperature of a coil 
is measured by the increase in resistance, or by resistance^ as 
it is usually stated, evidently the average temperature of the 
interior of the coil will be obtained, and this temperature 
rise will always be found to be greater than that which would 
be obtained by a thermometer on the outside, because the 
latter only gives the surface temperature, while the former 
gives the interior temperature. Where the temperature 
rise is specified by resistance, it is usually required that the 
rise shall not exceed 50° C, or the increase in resistance shall 
not exceed 20 per cent. This is a more rigid requirement 
than 40° C. by thermometer, unless the windings are very 
shallow. 

106, The American Institute of Electrical Engineers, 
in a report of the Standardization Committee, recommend 
specifically how these tests shall be made. A room temper- 
ature of 25° C. is recommended, and where this temperature 
varies, methods of correcting the results are given. The 
chief correction usually is for the temperature rise by 



Digitized by VjOOQ IC 



U DYNAMOS AND DYNAMO DESIGN §14 

resistance, when taken in a room whose temperature changes 
during the test. For instance, suppose a test is begun in 
the morning, when it is cool, and before it is completed the 
room temperature should rise 10° C, or 18° F. If the 
resulting temperature rise by resistance should be 60° C, 
obviously the correct rise that is due only to the heat devel- 
oped in the windings should be but 50° C. 

107. Where it is specified that the temperatures shall be 
taken after a continuous run at full load, it is meant that 
the machine shall be operated until it reaches its maximum 
temperatures. This can be told by observing the increase 
in resistance of the field windings, and, when these resist- 
ances become constant, it is pretty safe to assume that the 
machine as a whole has reached a constant temperature, 
and if it is then operated 25 per cent, longer time without 
increase of resistances, there will be little doubt but that the 
ultimate temperatures for continuous operation are reached. 

108. In regard to sparking, the other limit of the output 
of electric generators, it is usually required that machines 
shall operate at any load from no load to full load without 
change of position of the brushes and without appreciable 
sparking. Some customers require that machines shall 
operate from no load to 25 or 50 per cent, overload without 
injurious sparking. Generators intended for operating 
electric railways should be able to operate readily at 50 per 
cent, overload for periods of | hour without flasjiing or arc- 
ing at the brushes, and without glowiiig. Brushes in which 
the current density becomes too high will become red hot in 
spots, and this is called glo^wing. According to present 
practice, it is rot permissible to change the adjustment of 
the brushes at all under variable loads; they must be set 
once for all and maintained in that position. The correct 
position for the brushes is obtained by experiment when the 
generator is tested, and usually the position of the rocker- 
arm is indicated by a mark, so that, should the machine be 
taken apart for shipment or repairs, the correct position of 
the brushes may be readily obtained when reassembled. 



Digitized by VjOOQIC 



§14 DYNAMOS AND DYNAMO DESIGN 75 

109. In regard to the efficiency, it has been shown that 
it depends on the voltage of the generator to a considerable 
extent. It also varies somewhat with the speed, for a 
machine will have a proportionately greater output at a higher 
speed, but the losses do not usually increase quite as rapidly 
as the speed; hence, the efficiency is higher for machines 
operating at the higher speeds. It will be well to note in 
this connection that, according to the American Institute of 
Electrical Engineers, in the case of engine-type generators, 
it is not customary in computing the efficiency to include 
bearing friction and windage. This is because the bearing 
friction in this case rightfully should be considered a loss 
of the engine rather than of the dynamo, and the difficulty 
of determining the losses due to air friction or windage 
make it necessary to omit that also. Commutator brush 
friction, however, should be included in the calculations. 

no. Generators are required to be capable of develop- 
ing greater outputs than their normal ratings for a short 
time without excessive heating of any part and without 
serious sparking, so that in case of emergency they may be 
relied on for overloads. For electric-lighting service, gen- 
erators are usually required to stand an overload of 25 per 
cent, for from one to two hours, while for railway service an 
overload of 50 per cent, of like duration is frequently asked. 
Sometimes heavy overloads of long duration are demanded, 
but it is much better to depend on operation under full- 
load conditions rather than that on overloads, and this can 
always be done by making the capacity at full load ample* 
for the plant under consideration. 

Generators are usually required to be so wound as to 
develop a greater voltage on full load than on no load, but 
sometimes they are required to maintain the voltage con- 
stant at all loads. 

In either case, the machine must be provided with a com- 
pound field winding, for where only a shunt winding is 
used, the terminal voltage decreases as the load is increased, 
because of the loss of volts in the armature and commutator 



Digitized by VjOOQ IC 



76 DYNAMOS AND DYNAMO DESIGN §14 

due to their resistance, and also on account of the effects 
of armature reactions. For electric lighting, a rise of poten- 
tial of from nothing to 4 or 5 per cent, is usual, while for 
electric-railway and other power transmission a rise of 10 per 
cent, in voltage from no load to full load is customary. 

111. It is usually required that the insulation of the 
windings of a generator should meet two requirements: 
(1) the insulation must have at least a specified resistance 
between the windings and the frame; and (2) the insulation 
must withstand rupture when subjected to a high voltage 
applied between the windings and the frame. The usual 
requirement for insulation resistance is that for small 
machines it must exceed 1,000,000 ohms, or a megohm, as 
it is usually stated. For large machines, it is required that 
the insulation resistance be such that at the normal voltage 
of the machine the current that will pass through the insula- 
tion must not exceed uroirnnr P^^^ ^^ ^^^ full-load current. 
As an example, take a 100-kilowatt 125-volt generator. The 
full-load current is 800 amperes, and unriTnnr ^^ ^^^^ is 
.0008 ampere. Now, from Ohm*s law, 

^ = 7 = :S = i^^'^^« ^'^"^ 

The test for rupture of the insulation, or for dielectric 
strength, asi it is termed, is usually made with voltages at 
least 2J times as great as that at which the apparatus nor- 
mally operates. For small machines, for voltages under 800, 
from 1,000 to 1,500 volts is usually required, and for large 
generators, from 1,500 to 2,000 volts. The voltage is usually 
specified to be from an alternating-current source, as this is 
much more severe than a direct-current test. 

112. Supposing that the requirements that a certain 
dynamo is expected to fulfil are known, the generator should 
be tested in accordance with these requirements as nearly as 
possible. Having provided for the disposal of the electric 
power developed, the machine should be brought up to speed 
and the cold resistances of the field circuits and the room 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 



77 



temperature observed and recorded. The position of the 
brushes for sparkless commutation should be determined by 
experiment, varying the current from no load to full load. 
The correct voltages at no load should then be obtained by 
adjusting the shunt-field regulating rheostat and the load 
applied in equal steps, to determine if the rise in voltage 
with the load, or the compounding, is satisfactory. 



lisld MeguUUor 



113. If the generator does not compound sufficiently, 
the series-coils must be rewound with a greater number of 
turns. If the compounding is greater than desired, it may 
be reduced to any amount by shunting a part of the current 
around the series-coils. In Fig. 32 is shown a diagram of 
the connections for a 
compound-wound gener- 
ator in which the series- 
field coils terminate in a 
and d. It will be seen by 
the arrows that the cur- 
rent has two paths 
between these two points, 
one through the series- 
coil and the other through 
the resistance indicated 
by the zigzag line. This 
resistance is termed a 
shunt. The division of 
the current between 
these two paths depends 
on their relative resist- 
ances ; therefore, b y 
properly adjusting the resistance of this shunt, the current 
in the series-coil, and also the series ampere-turns, can be 
made anything desired. Such a shunt is usually made of a 
German silver strip, which is folded back and forth, so as 
to occupy as small a space as possible. In this connection, 
it should be remembered that the back ampere-turns of an 
armature are those that lie between the double angle of lead, 




Pig. 83 



Digitized by VjOOQ IC 



78 



DYNAMOS AND DYNAMO DESIGN 



§14 



and too great a lead of the brushes has the same effect 
as shunting a part of the current around the series-coils. 
Generators that do not require careful setting of the brushes 
to prevent sparking may sometimes be adjusted to com- 
pound properly by changing the position of the brushes. 

114. When all is adjusted as desired, the test to deter- 
mine the temperature rise under full load, often termed the 
heat test, is begun. After this is completed, the tempera- 
tures of the various parts are taken, the resistances of the 
field coils measured, and the compounding is again taken 
to determine if everything is satisfactory and within the 
requirements. 

115. While the machine is still hot, it should be sub- 
jected to the insulation tests already described. To measure 
the insulation resistance, it is necessary to use a direct cur- 
rent, and the most convenient instrument to use is the ordi- 
nary voltmeter. The connections for this test are shown in 




Pio. 88 

Fig. 33, in which the insulation resistance of an armature 
is being measured by means of a 600-voIt voltmeter on a 
500-volt direct current. One terminal of the voltmeter con- 
nects to the line and the other to the shaft of the armature, 
while from the commutator a wire connects to the other 
side of the line. The current from the line passes to the 



Digitized by VjOOQ IC 



§14 DYNAMOS AND DYNAMO DESIGN 79 

winding through the commutator, thence leaks through the 
insulation to the core, and out on the shaft to the voltmeter, 
and back to the line. It is clear that what current passes 
through the insulation also passes through the voltmeter and 
causes it to indicate on the scale the voltage existing at its 
terminals. As an illustration of the method, suppose the 
voltmeter used in a certain case has a resistance of 
60,000 ohms, and that it indicates 10 volts when connected, 
as shown in Fig. 33, on a oOO-volt line. If there is 10 volts 
difference of potential at the voltmeter terminals, there 
must be 490 volts between the armature winding and the 
core, because there are 500 volts altogether, and the cur- 
rent is so small that there is no appreciable drop in con- 
nections. The same current passes through the insulation 
as passes through the voltmeter ; hence, the resistances of 
the voltmeter and insulation are in the same proportion as 
the voltages across them. The voltage across the insulation 
is 49 times that of the voltmeter, and the insulation resist- 
ance is 49 X 60,000 = 2,940,000 ohms, which is nearly 
3 megohms. The resistance of a voltmeter is usually given 
with the instrument. It will be apparent, after a little 
thought, that a voltmeter with a very high resistance is 
preferable. It is also advisable that the line voltage be not 
less than 500 volts. An alternating current and an alterna- 
ting voltmeter cannot be used for measuring the insulation 
resistance, since the armature will act as an electric con- 
denser to some extent, and an alternating current can pass 
through a condenser, while a direct current cannot. 

116, The test for dielectric strength, as has already been 
stated, is made with an alternating current. One terminal 
of the high-voltage line is connected to the windings, and 
the other terminal is connected to the frame of the machine. 
In case of defective insulation, the circuit will be completed 
and an excessive current will flow that may be indicated by 
the blowing of a fuse or other indicating device. In this 
case it is not necessary to measure the actual current flowing, 
since the test is only to show the ability of the insulation 



Digitized by VjOOQ IC 



80 



DYNAMOS AND DYNAMO DESIGN 



§14 



to resist puncture. As a matter of fact, this breakdown 
test is by far the more important of the two, for it is pos- 
sible that a weak spot in the insulation may be protected by 
a film of varnish and show a high insulation resistance, but 
a severe test on a high voltage will reveal the defect 
promptly. 

11 7, The preceding tests are those to which a dynamo 
is subjected to determine its adaptability for certain 
requirements. There are, of course, very many other tests 
to which generators are subjected for determining certain 




jor\ 




PIO. 84 

particulars, but these are not, as a rule, very important in 
comparison to those just discussed. Among the more impor- 
tant tests is that from which a magnetization curve is 
obtained experimentally. To make this test, the connec- 
tions are made according to Fig. 34. The dynamo is sep- 
arately excited, so that the field current can be readily 
changed. In the field circuit is placed an ammeter A and 
adjustable resistance R^ and a voltmeter V is connected 
across the brushes. The speed of the generator should be 
maintained constant, if possible, during the test, and the 



Digitized by VjOOQ IC 



§ 14 DYNAMOS AND DY:NAM0 DESIGN 81 

brushes should be placed in the neutral position, or in such 
a position that the voltage obtained is the greatest under 
the conditions of speed and field current. The field current 
should now be varied by means of the rheostat R, and caused 
to take a series of values, the amperes in each case being 
read from the meter A^ and the volts between the brushes 
read from the meter V. It is best to take the average of 
the readings obtained both with increasing and decreasing 
field currents, because on decreasing the field current, the 
iron of the magnetic circuit tends to retain its magnetism, 
and the voltages obtained are usually greater than those 
obtained with increasing currents. From the amperes in 
the field circuit, and the number of turns on the field coils, 
the various field ampere-turns can be computed, while the 
total flux per pole can be computed from the known number 
of armature conductors, the speed and terminal voltage. If 
these two values are plotted on cross-section paper, a mag- 
netization curve like that shown in Fig. 11 will be obtained. 
A magnetization curve is of importance in computing the 
field windings. Where it is desired to rewind a machine for 
another voltage, it is a comparatively simple matter if a 
magnetization curve is obtained from the frame before the 
original windings are removed. The winding calculations 
are made after the manner already explained in connection 
with the design. 



44—17 



Digitized by VjOOQIC 



Digitized by VjOOQIC 



DIRECT-CURRENT MOTORS 



PBINCIPLES OF OPERATION 



DITN^AMOS ANB MOTORS COMPARED 

!• A dynamo may be defined as a machine for the gen- 
eration of an electromotive force and current by the motion 
of conductors through a magnetic field. This motion and 
the force necessary to maintain it must be supplied by a 
steam engine or other source of power. On the other hand, 
a motor may be defined as a machine for supplying mechan- 
ical power when supplied with an electric current from some 
outside source. The motion and the force necessary to 
maintain it are in this case supplied by the reaction between 
the current flowing in a set of conductors and the magnetic 
field in which the conductors are placed. 

2. As far as the electrical features of a direct-current 
motor are concerned, they are almost identical with those 
of the continuous-current dynamo. The differences in the 
two that occur in practice are very largely differences 
in mechanical details that are necessary to adapt the motor 
to the special work that it has to do. No matter what the 
mechanical design of motors may be, they all consist of the 
same essential parts as the dynamo, namely, field magnet 
and armature with its commutator, brush holders, etc. 

§15 

For notice of copyright, see page immediately following the title page. 



Digitized by VjOOQ IC 



DIRECT-CURRENT MOTORS 



§15 




ACTION OF MOTOR 

3. It is necessary to consider carefully the forces acting 
in a motor in order to understand clearly the behavior of 

different kinds of 
motors when oper- 
ated under given con- 
ditions. In order to 
do this, we will con- 
sider the force acting 
on a conductor that 
^o- * is carrying a current 

across a magnetic field. Suppose the arrows, Fig. 1, repre- 
sent magnetic lines of force flowing between the pole faces 
of the magnet A^, 5, and let a represent the cross-section of 
a wire lying at right 
angles to the lines. 
So long as no current 
flows through the 
wire, the field will not 
be distorted, and 
there will be no ten- 
dency for the wire to fio. 9 
move. If the ends of the wire are connected to a battery so 
that a current flows down, this current will tend to set up 
lines of force around the wire, as shown by the dotted 

circles, Fig. 2. It will 
be noticed that these 
lines tend to oppose 
the original field be- 
low the wire and 
make it more dense 
above the wire. The 
resultant effect is 
that the field is distorted, as shown in Fig. 3, and the wire 
is forced downwards. 

4, The action described in the simple case just given is 
essentially that which takes place in an electric motor. 





Pio. 8 



Digitized by VjOOQ IC 



§ 16 DIRECT-CURRENT MOTORS 8 

The magnet is excited by means (rf current taken from the 
mains to which the motor is connected. Current from the 
line is led into the armature windings by means of the com- 
mutator and brushes, and reacts on the field, thus driving 
the armature around. 

6, By referring to Fig. 3, it will be seen that in a motor 
the conductors are forced across the field by the reaction of 
the armature current on the field; that is, the force exerted 
by the magnetic field on the armature conductors of a motor 
is in the same direction as the motion of the armature. 
This force is made use of for doing mechanical work. The 
armature of a dynamo is driven by means of a steam engine 
or other source of power, and the armature conductors are 
made to cut across the magnetic field, this motion causing 
the generation of an E. M. F. When the outside circuit is 
closed, so that current flows through the armature con- 
ductors, this current reacts on the field in such a way as to 
oppose the motion of the armature. The more current the 
dynamo supplies, the greater is the drag between armature 
and field and the more work the steam engine has to do to 
keep the dynamo operating. In the case of a motor, the 
greater the load applied to the pulley, the greater must be 
the twisting action between the armature and field to keep 
up the motion, and the greater the amount of current that 
must be supplied from the line. It is thus seen that as 
regards the twisting action between the armature and field, 
the motor is just the opposite of the dynamo, the action 
in the former case being with the direction of motion and in 
th^ latter case against it. 



COUNTER E. M. P. OP MOTOR 

6. Whenever a conductor is moved in a magnetic field 
so as to cut lines of force, an E. M. F. is induced in the con- 
ductor. In the case of a dynamo this E. M. F. is made use of 
to set up currents in outside circuits. In other words, the 
E. M. F. is the cause of the flow of current, and conse- 
quently the E. M. F. is in the same direction as the current. 



Digitized by VjOOQtC 



4 DIRECT-CURRENT MOTORS §16 

In a motor we have all the conditions necessary for the 
generation of an E. M. F. in the armature. It is true that 
the armature is not driven by a belt as in the case of a 
dynamo, but by the reaction between the field and armature. 
This, however, makes no difference so far as the generation 
of an E. M. F. is concerned. 

When a motor is in operation, there must be an E. M. F. 
generated in its armature, and for the present we shall term 
it the motor E. M. F. Take the simple case shown in 
Fig. 3; as the conductor is forced down, it will pass across 
the magnetic field, and an E. M. F. will be induced in it. 
Also, by applying the rule for determining the direction of 
the induced E. M. F., we see that it must be directed 
upwards, that is, toward us along the conductor (the direc- 
tion of motion being down and the direction of the field 
from left to right). The current flowing in the conductor 
is flowing away from us, or is being opposed by the E. M. F. 
We may state, then, in an electric motor the E. M. F, gen- 
erated in the armature is opposed to the current flowing 
through the armature, 

7. Owing to the fact that the motor E. M. F. is opposed 
to the current, it is commonly spoken of as the counter 
E. M. F. of the motor. It is important that the student 
should clearly understand the generation of this counter 
E. M. F. and its relation to the current. 

When dynamos were first operated as motors, the exist- 
ence of this counter E. M. F. was thought to be a draw- 
back because it tended to keep the current out of the motor. 
It was soon found, however, that the counter E. M. F. was 
essential to the operation of the motor. In a dynamo we 
have thecounter torque or drag on the armature, which the 
engine has to overcome ; in the motor the torque assists the 
motion, and we have the counter E. M. F., which is opposed 
to the line E. M. F. Suppose a motor is connected to a 
constant-potential dynamo and that the armature of the 
motor is held from turning. There will then be no counter 
E. M. F. because the conductors are not cutting across the 



Digitized by VjOOQ IC 



§15 



DIRECT-CURRENT MOTORS 




field. The current that will /low through the armature will 
be fixed by Ohm*s law, and will be equal to the applied 
E. M. F. divided by the resistance of the armature. Since 
the resistance of the ai mature is usually very low, the cur- 
rent will be quite large. Now, it will be noted that the 
motor is doing no useful work because the armature is not 
turning and all the energy supplied is expended in heating 
the armature. If the armature is released, it will at once 
run up to speed and at the same time the current will 
decrease. If a brake is put on the pulley, it will be found 
that as the load is increased by tightening the brake, the 
current through the armature increases, but it is still much 
smaller than when the arma- 
ture was held from turning. 
The ohmic resistance is the 
same whether the arma- 
ture is in motion or not, so 
that the large decrease in 
current is not due to an 
increase in resistance. 
Moreover, the dynamo sup- 
plying the current main- 
tains a practically constant 
E. M. F., and the reason 
that the current decreases 
is that the moving armature 
introduces a counter 
E. M. F. into the circuit, 
and the E. M. F. that is 
effective in forcing cur- 
rent through the motor 
armature is the difference 
between the applied E. M. F. and the counter E. M. F. 
Suppose the constant-potential dynamo A, Fig. 4, sup- 
plies current to the motor B. The pressure required to 
force a given current / through the armature is, from Ohm's 
law, IRai where ^„ is the resistance of the armature. The 
total impressed E. M. F. £", i. e., the line E. M. F., must 




PIO. 4 



Digitized by VjOOQ IC 



e DIRECT-CURRENT MOTORS §15 

be equal to the counter E. M. F. plus the E. M. F. required 
to overcome the armature resistance. Hence, if E^ is the 
counter E. M. F. 6i the motor, we have 

£=£^ + /R, (1) 
or E^ = E--IK (2) 

The counter E. M. F. cannot be measured directly, but its 
value can be calculated if we know E, /, and R^. 

Example. — A motor armature has a resistance of .05 ohm, and when 
carrying a certain load requires 200 amperes in its armature. The line 
pressure is 220 volts. What is the counter E. M. F. at this particular 
load? 

Solution. — The pressure taken up in overcoming the armature 
resistance must be 200 X .05 = 10 volts; hence, the counter E. M. F. 
must be jE;» = 220 — 10 = 210 volts. Ans. 

8. It is evident from formula 2 that if the current 
becomes very small (in which case the load in the motor 
must be very light), the counter E. M. F. E^ becomes 
nearly equal to the line E. M. F. E. If E and E^ were 
exactly equal, no current would flow through the armature. 
The product of E and /represents the power supplied from 
the line. Even if no load is applied at the pulley, some 
power must be supplied to keep the armature in motion; 
consequently, the motor must take some current whether it 
is loaded or not, and as current cannot flow into the motor 
unless the counter E. M. F. is less than 
the applied E. M. F., it follows that the 
counter E. M. F. can never be quite as 
large as the applied, although it may be . 
very nearly so at light loads. 

The fact that a motor when running 
tends to set up an E. M. F. opposed to 
the line E. M. F. is easily shown. In 
Fig. 5, A is the armature of a motor, m an 
ammeter, r a resistance, and i, ^, and 3^ 
switches. Suppose switches 1 and 2 are closed and switch 3 
open, and that the armature is running at full speed. Then 
open switches 1 and ^, and at once close switch 3, The 




Digitized by VjOOQ IC 



§15 DIRECT-CURRENT MOTORS 7 

ammeter m will show that a current is flowing, and a 
current will continue to flow until the energy stored in the 
revolving armature is used up. This generation of current 
by the armature soon brings it to a stop. In fact, allowing 
the motor to act as a generator in this way makes a very 
effective electric brake, and this method of braking is used 
quite largely for electric elevators and also, to some extent, 
for street cars. The ammeter will also indicate that the 
current flows as shown by the dotted arrows; whereas, 
when the motor was running from the line, the current 
flowed as shown by the full-line arrows. In other words, 
the current set up by the motor E. M. F. is opposite to that 
set up by the line E. M. F. 

9. In a dynamo, the electrical energy developed in the 
armature windings is the product of the current and the 
E. M. F. generated in the windings. Most of this energy, 
but not all, is available at the terminals of the dynamo. 
Some of the electrical energy is lost in forcing the current 
through the windings. In the motor, a certain amount of 
electrical energy is supplied, and part of this is converted 
into mechanical energy within the armature. This part is 
represented by the product of the counter E. M. F. and the 
current, and it is evident that the higher the counter E. M. F. 
compared with the applied E. M. F., the more efficient is 
the motor. Action and reaction are always equal and oppo- 
site. In the dynamo, current is supplied to an outside cir- 
cuit, and the reaction appears as the counter torque or the 
drag that the engine has to overcome. In the motor, 
mechanical energy is delivered at the pulley, and the reac- 
ticn appears as the counter E. M. F. that the line E. M. F. 
has to overcome. This point is dwelt on here, because stu- 
dents when first studying the principle of electric motors 
seem to have difficulty in understanding the action of this 
counter E. M. F. They think it is something that prevents 
the proper action of the machine, whereas, it does not do 
so any more than the drag that the steam engine has to 
overcome prevents the generation of current by a dynamo. 



Digitized by VjOOQ IC 



DIRECT-CURRENT MOTORS § 15 



MOTOR BFPICIENCT 

10. There are certain unavoidable losses in a motor just 
as in a dynamo and the losses are of the same character. 
There are the electrical losses, consisting of the PR loss 
in the field, armature, and commutator. There are the core 
losses (hysteresis and eddy currents), together with the 
usual mechanical losses, such as friction at the bearings, 
brush friction, and windage. As in the dynamo, there are 
three eflSciencies to be considered: Xh^ commercial efficiency^ 
the electrical efficiency^ and the efficiency of conversion. 

11. Of these three the commercial efficiency is the 

most important, because it shows the percentage of the total 
power supplied that is converted into useful power at the 
pulley. The commercial efficiency takes account of all 
the losses in the motor. In other words, if W^ represents 
the total watts supplied from the lines, and IV the watts 
available at the pulley, then 

W 
commercial efficiency f/ = tf^ (3) 

12. The electrical efficiency shows the relation be- 
tween the amount of power developed within the armature 
and the total amount of electrical power supplied. The 
amount of power developed within the armature is equal 
to the counter E. M. F. multiplied by the current, and it is 
equal to the electrical power supplied less the various elec- 
trical losses. If IF, is tlfc amount of power developed in 
the armature, then 

electrical efficiency 17^ = -rr^ (4) 

13. All the power developed in the armature is, how- 
ever, npt available at the pulley. A part of it is used in 
overcoming the core losses and friction of the various parts. 
The efficiency of conversion is a measure of the ability of 
the motor to convert the electrical power in the armature 



Digitized by VjOOQ IC 



§15 DIRECT-CURRENT MOTORS 9 

into mechanical power delivered at the pulley. The power 
delivered at the pulley is IV ; hence, 

JV 
efficiency of conversion [/^ = -rp (6) 

14, The student should compare these efficiencies with 
the corresponding efficiencies for the dynamo. Since 

W W IV 

^X Trr= TTF, it follows that the electrical efficiency mul- 
tiplied by the efficiency of conversion gives the commercial 
efficiency as with the dynamo. 

From the foregoing it will be seen that the points that are 
in favor of an efficient dynamo are also in favor of an 
efficient motor. The armature should be of low resistance, 
the shunt fields of high resistance, the core losses should be 
made as low as possible, and all friction should be kept 
within reasonable limits. In short, so far as electrical fea- 
tures are concerned, the machine that makes a good dynamo 
will also make a good motor, though its mechanical features 
may not be particularly adapted for some kinds of motor 
work. 

16. The various efficiencies of a motor, as with a 
dynamo, vary with the load because the current varies with 
the load. At no load the current is very small, and all the 
energy taken from the mains goes to supply the losses. No 
useful power is delivered at the pulley, and the commercial 
efficiency at no load is therefore zero. As the load is 
increased, the commercial efficiency rises until the maximum 
is reached at a point depending on the design of the 
machine, as in the case of a dynamo. If the load on the 
motor is forced too high, the /* R loss in the armature 
becomes excessive and the efficiency falls off. 

16. Current Required by Motors, — The current that 
any given motor will take from the line at full load will be 
equal to the total number of watts supplied, divided by the 
line voltage. The input will be equal to the output divided 



Digitized by VjOOQ IC 



10 



DIRECT-CURRENT MOTORS 



§15 



TABIiB I 

COMMERCIAI. EFFICIENCY OF MOTORS 



Output Brake Horse- 
power at Full Load 


Commercial Efficiency 


I 


.650 


5 


•750 


10 


.820 


.2i 


.850 


15 


.860 


25 


.880 


50 


.890 


75 


.900 


100 


.910 


150 


.920 


200 


•9*5 



by the commercial efficiency. The full-load current, there- 
fore, depends on the efficiency of the motor, and hence may 
differ by a limited amount in motors of different design. A 
motor might be designed to give a high efficiency at fairly 
light loads with a less efficiency at full load. Or it might be 
designed for a high efficiency at full load, with lower 
efficiencies at light loads. In any event the current corre- 
sponding to any brake horsepower output will be 

H. P. X 746 



/ = 



(6) 



where 



Ey.U 

= current taken from line; 

= brake horsepower (horsepower delivered 

at pulley); 
= line E. M. F. ; 
= commercial efficiency corresponding to 

given load. 

Example. — A motor is capable of delivering 10 brake horsepower 
when fully loaded, and its full-load commercial efficiency is 85 per cent. 
What full-load current will this motor take from 220-volt mains ? 



/ 
H. P. 

E 
U 



Digitized by VjOOQ IC 



§15 DIRECT-CURRENT MOTORS 11 

Solution.— In formula 6 we have H. P. = 10, E = 220, ^= .85; 
hence, 

It will be noticed in Table I that the efficiency at first 
increases quite rapidly with increase in size, but as the 
output becomes greater, the increase in efficiency becomes 
smaller. 

17. The current that a given motor takes at full load is 
usually marked on the name plate of the motor. If it is not 
given, the foregoing values of the commercial efficiency may 
be used for approximate calculations of the current. It 
must be understood, however, that these values of the effi- 
ciency are by no means fixed and might easily vary 2 or 
3 per cent, either way from the values given. 



TORQUE 

18. The torque of a motor is the twisting or turning 
effort exerted on the armature. Suppose the pulley of a 
small motor is grasped by the hand and current sent 
through the armature while the field is excited. It will be 
found that there is a strong twisting action, or tendency for 
the pulley to turn, and this twisting action is known as the 
torque. Suppose we have a motor arranged as shown in 
Fig. 6. P is the pulley on which the shoes j9, B press; the 
pressure between the shoes and the pulley can be adjusted 
by means of the thumbscrews A^, N, To the lower block is 
attached an arm A provided with the point or knife edge C. 
This point presses on the platform scales 5, so that the 
pressure exerted by the arm can be readily measured. The 
arm is balanced by the weights IV, so that when the motor 
is standing still the arm A is just counterbalanced and there 
is no pressure between it and the scales. A device of this 
kind is known as a Prony brake, and by means of it the 
power delivered at the pulley of the motor can be measured. 



Digitized by VjOOQ IC 



12 DIRECT-CURRENT MOTORS §15 

Suppose for the present that the blocks B are clamped so 
tightly that the armature cannot turn when the normal 
full-load current of the motor is sent through it. Although 
the pulley cannot turn, there will be a strong torque, or 
tendency to turn, and the pin C will be pressed down against 
the scale platform, the scale beam registering the number 
of pounds pressure. It is evident that pressure obtained on 
the scale will depend on the length of the radius R from the 
center of the pulley to point C, so that it would mean noth- 
ing to state that the torque was so many pounds unless the 
length of the radius at which this reading was taken was 



FlO. 6 

also stated. For example, if the radius were made one-half 
as great, it is easily seen that the pressure would be doubled. 
In other words, torque must be expressed as a moment, 
i. e., by the product of a force into a lever arm. Torque is 
therefore expressed \x\ pound feet ox foot-pounds^ the former 
being preferred because the foot-pound is more commonly 
used for the unit of work. For a given current in the arma- 
ture, and given field strength, the product of force X lever 
arm, i. e., the torque, is a constant quantity. The number 
of pound feet will be the same no matter what lever arm is 
used, because the larger the arm, the less the force exerted 



Digitized by VjOOQ IC 



§16 DIRECT-CURRENT MOTORS 18 

at the end of it. The number of pounds pressure on the 
scale will, however, depend on the length of the arm A, and 
the student should distinguish carefully between the force 
exerted in pounds, and the torque in pound feet. A motor 
might be giving a very small torque and yet exert many 
pounds belt pull, provided the pull is exerted at the end of 
a very short lever arm. 

19. The student must also carefully distinguish between 
torque and power. Power is the rate at which work is 
done. In Fig. 6, if the brake is clamped so that the arma- 
ture cannot turn when a current is sent through it, a torque 
will be exerted, but the motor is delivering no power 
because it is not running. It is possible, therefore, to have 
torque without power. The torque is simply the turning 
moment exerted on the armature, and in order that power 
may be delivered, the armature must be allowed to revolve. 

20. Suppose, in Fig. 6, that the brake is loosened enough 
to allow the armature to turn but yet keep a considerable 
amount of friction between the shoes and the pulley. A 
pressure will be exerted on the scale as before, and this 
pressure will depend on the torque exerted by the motor. 
If Pis the pressure in pounds on the scale, / the tangential 
force exerted at the pulley rim, and r the radius of the 
pulley, the torque is PR, which must also be equal to 
the force exerted at the pulley rim multiplied by the radius 
of the pulley; that is, 

torque = PR = / r 

Now, there is a steady frictional resistance of / pounds 
at the rim of the pulley; or we can look at it as if the 
motor were continuously winding up a cord to which is 
attached a weight of / pounds. During each revolution 
the weight is lifted or the resistance overcome through a 
distance equal to the circumference of the pulley; that is, 
the work done in one revolution is 2 jt r/ foot-pounds, where 
the radius of the pulley r is expressed in feet. Now, if the 
motor is running S revolutions per minute, the number of 



Digitized by VjOOQ IC 



14 DIRECT-CURRENT MOTORS §15 

foot-pounds of work done per minute will he % it r p S, But 
the amount of work done per minute is a measure of the 
power developed by the motor, and rp is the torque exerted, 
so that the power is equal to 2^ 7^5, where T is the torque 
in pound feet. We may, therefore, write the general 
relation 

power = 2w X torque X speed (7) 

Since the quantity 2^: is constant, it follows that the 
power developed depends on the torque and on the speed, 
and while torque and power are related as shown above, they 
are by no means the same thing. 

21. Sometimes the torque of a motor at a given current 
is spoken of as so many pounds. When this is done, it is 
always understood that this number of pounds pull is exerted 
at the end of a 1-foot radius, in which case the number of 
pounds pull is numerically equal to the torque in pound 
feet. It is always better, however, to express torque in 
pound feet so that there will be no confusion. 

22. From formula 7 it is easily seen that a motor can 
be designed to give a certain amount of power by using 
different values of the torque and speed. For example, sup- 
pose a motor were to have a capacity of 10 horsepower. 
This capacity could be obtained by making a motor that 
would give a very strong torque and run at a low speed, or 
the same power could be obtained with a motor giving a 
small torque and running at a high speed. Some classes 
of work demand a strong torque and low speed, while in 
other cases a light torque and high speed are desired. The 
matter of torque is therefore an important one, and has to 
be carefully considered when motors are being designed or 
selected to do a given class of work. 

In formula 7 the power is given in foot-pounds per min- 
ute, because the torque T is expressed in pound feet, and 
the speed in revolutions per minute. Since 1 horsepower is 
equivalent to 33,000 foot-pounds per minute, we have 



Digitized by VjOOQ IC 



816 DIRECT-CURRENT MOTORS 16 

H. P. = ^ ^ ^}.f]l^ ^^ = .0001904 7-5 (8) 

By making a test with the Prony brake, as shown in 
Fig. 6, the torque T is easily obtained, because it is equal 
to the scale reading multiplied by the lever arm R. The 
speed 5 can be measured with a speed counter, so that the 
horsepower can be at once calculated from formula 8 
Formula 8 may be changed as follows : 

^ _ 33,000 X H. P. _ H. P. ^ 

- 2 X 3' 1416 T "" .0001904 T ^^ 

which gives the speed at which a motor must run in order 
to deliver a given horsepower at a given torque. 

Or, ^ ^ 33,000 X H. P. _ H. P. ^^ 

2 X 3.1416 5 ■" .0001904 5 ^ ^ 

which gives the torque corresponding to a given horsepower 
and speed. 

Example. — A given motor is designed for an output of 10 horse- 
power, and is run on a 280-volt constant-potential circuit. When dri- 
ving a certain piece of machinery it requires an electrical input of 
85 amperes at 230 volts. It is desired to find the actual horsepower 
required to drive this machinery. The motor is disconnected from its 
load and a Prony brake rigged up as shown in Fig. 6. The thumb 
nuts are screwed up until an ammeter in the motor circuit indicates 
85 amperes, the pressure across the circuit being 280 volts. Under 
these conditions the pressure on the scale plat^rm is found to be 
24 pounds, and the speed of the motor 800 revolutions per minute. 
The horizontal distance between the center of the shaft and the point 
pressing on the scales is 30 inches, (a) What is the horsepower output 
of the motor, and (d) what is the commercial efficiency of the machine ? 

Solution. —(^1) The distance ^, Fig. 6, is 80 in. = 2^ ft. The 
pressure on the scales is 24 lb. ; hence, the torque = 7" = 24 X 2^ 
= 60 lb. ft. S = 800; hence, substituting in formula 8, we have 

„„ 2x8.1416x60x800 « ^ , ,^ ^ 

H. P. = —^ ^ g^ = 9.14 H. P., approximately Ans. 

(d) The commercial efficiency is the ratio of the output to the input. 
The input is 35 X 230 = 8,050 watts. The output, since there are 
746 watts in a horsepower, is 9.14 x 746 = 6,818.4 watts. The commer- 

cial efficiency is therefore ' ^^ = .847, or 84.7 per cent. Ans. 
44—18 



Digitized by VjOOQ IC 



16 DIRECT-CURRENT MOTORS §16 

23. Relation Between Torque and Current. — Sup- 
pose we have a motor in which the friction and core losses 
are negligible. The power delivered at the pulley would 
then be equal to the power developed in the arma- 
ture. The power in watts delivered at the pulley is 

2X3.1416X^X5X746 u t - ^u ^ • a 
QQ~AAA » where T is the torque m pound 

feet. If E^ is the counter E. M. P., this E. M. F. must be 

equal to — ^ ,^, (see Dynamos and Dynamo Desis[n. 

^ ;;/ X 10' X 60 ^ -^ '^ ^ ' 

Part 2). In this formula / is the number of poles, ^ the 

flux to or from each pole, Z the total number of armature 

conductors, m the number of paths in the winding, and 5 

the speed in revolutions per minute. The power developed 

in the armature is equal to E^ X /, where / is the current 

supplied to the armature. Neglecting losses, the power 

developed in the armature is equal to the power delivered 

at the pulley, or 

2X 3.1416 X rx5x 746 __ p^ZS j 

33,000 "■ /« X 10' X 60 ^ 

The speed 5 cancels out from each side of this equation, 
and solving for T we have 

J. dd,000p ^ZI 

- 2 X 3.1416 X 746 X mx 10' X 60 

_ .00117 p Z ^ I ,^ ^ 

"" mx 10- ^ ^ 

XT r • i_ . .^^WtpZ. ^ ^ 
Now, for any given motor the quantity ^-r- is fixed 

in value, because the number of poles/, the number of con- 
ductors Z, and the number of paths m cannot be changed. 
For a given motor, then, the torque T can be changed by 
changing the field flux ^ or the armature current /, and for 
a giveo field flux the torque depends on the current and is 
independent of the speed. In some kinds of motors the 
field flux is dependent indirectly on the speed, so that 
the torque varies with the speed. An examination of 
formula 11 shows that if a motor is to exert a strong 



Digitized by VjOOQ IC 



§16 



DIRECT-CURRENT MOTORS 



17 



torque with a given current /, it must be provided with a 
large number of armature conductors and also have a strong 
field. 

Example. — A four-pole motor armature has 240 conductors arranged 
in a four-path winding. The flux from each pole is 2,000,000 lines. 
(a) What torque will be exerted when a current of 100 amperes is sup- 
plied to the armature ? (d) Neglecting losses, what would be the pull 
at the rim of the pulley if the pulley were 20 inches in diameter ? 

Solution.— (a) In this case, / = 4, Z = 240, w = 4, * = 2,000,000. 
and / = 100. Applying formula 11, we have 



T = 



.00117 X 4 X 240 X 2,000,000 X 100 
4X1,000.000 



= 56.16 lb. ft. Ans. 



(d) The radius of the pulley is ^ X fj = f ft. ; hence, the pull at the 

Ans. 



rim must be ^^il? = 67.39 lb. 



ABMATUBB REACTION 

34. Armature reaction is present in motors as in 
dynamos, but its effects are somewhat different. Lret Fig. 7 




PIO. 7 



represent the pole pieces and armature of a two-pole motor, 
and suppose current to be sent into the armature so that it 
flows, say, downwards in the right-hand conductors and 



Digitized by VjOOQIC 



18 



DIRECT-CURRENT MOTORS 



§15 



upwards in those on the left. Suppose for the present that 
the brushes are directly on the neutral line midway between 
the poles, as shown. The effect of the armature currents 
will be to cross-magnetize the field, as shown by the dotted 
lines, and by considering the direction of the cross-mag- 
netism as related to the magnetism set up by the field mag- 
net, it will be seen that the resultant effect is to weaken the 
pole corners b^ b and to strengthen ^, a. It is also evident 
that the direction of rotation will be as shown by the arrow, 
the effect of the cross-magnetization being to shift the field 
backwards as regards the direction of rotation instead of 
forwards, as in the case of a dynamo. 

25. Fig. 8 shows the same armature with the brushes 
shifted back to the non-sparking point. The shifting of the 
brushes brings into play the back ampere-turns that are 




PlO. 8 



included between the double angle of lead. It will also be 
seen that these back ampere-turns tend to demagnetize the 
field, as shown by the dotted arrow at a, the action of the 
armature in this respect being the same as the action in a 
dynamo. It may also be noted here that if it were possible 
to operate the motor without sparking, with a forward lead 
of the brushes, the back ampere-turns would tend to mag- 
netize the field. It is important that motors be designed so 



Digitized by VjOOQ IC 



§15 DIRECT-CURRENT MOTORS 1ft 

that the shifting of the sparking point from no load to full 
load shall be small. This means that the field should be 
stiffs or powerful, and the effects of armature reaction made 
as small as possible by adopting the various methods 
explained in connection with dynamo design. In modern 
motors of good design the shifting of the neutral point is 
very slight. Carbon brushes are used almost exclusively, 
and they may be left in the same position from no load to 
full load without sparking. 

26. It is instructive to note, in connection with motor 
armature reaction, that if the brushes have any lead for- 
wards or backwards and a current is sent through the 
armature alone, the field being unexcited, the armature will 
revolve, because the armature reaction will set up a field for 
the armature currents to react on. The torque produced 
would, of course, be very small, because the field set up in 
this way would be very weak. 



CLASSES OF MOTORS 

Vt. Direct-current motors, like dynamos, are generally 
classed according to the methods adopted for exciting the 
field magnets. This naturally divides motors into the fol- 
lowing classes : iX) Shunt-wound; {^) series-wound; (d) com' 
pound'ivound. 

Motors may also be operated with their fields supplied 
from one source of current and their armatures from 
another. By far the larger part of the motors in use 
belong to the first two classes, the third class being used 
only to a limited extent. Nearly all motors are operated on 
constant-potential circuits. The voltage across the termi- 
nals is maintained constant or nearly so by the dynamo sup- 
plying the system, and the current taken by the motor 
varies with the load. At one time motors were operated on 
constant-current arc-light circuits to a limited extent. In 
this case the current through the motor remains constant, 
and the voltage across its terminals increases with the load. 



Digitized by VjOOQ IC 



20 DIRECT-CURRENT MOTORS §16 



SHUNT MOTORS 

38. The sliunt-Tround motor is identical, so far as 
its electrical construction is concerned, with the shunt- 
wound dynamo. 

These motors are operated on con- 
stant-potential systems, the motor 
SI being connected directly across the 
X"^!?^ P mains when running, as shown in 
Fig. 9, where A is the armature and F 
is the field. If £, the E. M. F. be- 
tween the mains, is maintained con- 
stant, the current flowing through the 
shunt field will be constant. The field 
Pw- • coils will therefore supply approxi- 

mately the same magnetizing force, no matter what cur- 
rent the armature may be taking from the mains. The 
strength of field would be practically constant if there were 
no demagnetizing action of the armature. 




ACTION OF SHUNT MOTOR 

29. When a load is applied, the motor must take suffi- 
cient current to enable the armature to produce a torque 
large enough to carry the load. In order to allow this cur- 
rent to flow, the counter E. M. F. must lower slightly, and 
as the field is nearly constant, this means a slight lowering 
of speed. At the same time it must be remembered that 
the back ampere-turns will make the field flux slightly less 
when the motor is loaded than when it is not loaded, and 
this weakening of the field tends to keep the speed up. The 
net result, therefore, is that a shunt motor operated on a 
constant-potential circuit falls off slightly in speed as the 
load is applied, but if the motor is well designed and has a 
low-resistance armature, the falling off in speed from no 
load to full load will be very small. It is this speed-regu- 
lating feature that makes the shunt motor so widely used. 
If the load should be accidentally thrown off, there is no 



Digitized by VjOOQ IC 



§15 DIRECT-CURRENT MOTORS 31 

tendency to race, and the motor automatically adjusts itself 
to changes in load without materially changing its speed 
and without the aid of any mechanical regulating devices. 
Whenever the load on a shunt motor is changed, the cur- 
rent taken from the mains also changes, because the speed 
and field strength remain practically constant. Now, we 
have already seen that the counter E. M. F, £^ = E — I K„, 
where E is the line E. M. F., R^ the armature resistance, 
and / the current. From this we have 

7=£^ (13) 

Now, the line E. M. F. E is constant, and the armature 
resistance R^ 's also constant, so that if / is to increase with 
increase in load, it follows that E^ must decrease. Since 
the armature resistance is very low, a small falling off in 
the counter E. M. F. is sufficient to allow the increase in 
current needed to furnish the larger torque required to 
carry the added load. If the armature resistance were 
high, a considerable falling off in speed would accompany 
an increase of load. 



SPEBD REGULATION OF SHUNT MOTORS 

30. In any direct-current motor we have the relation 



£« = 



w X 10' X 60 



C ^ X 10" X 60 X E^ .^«. 

or S = —y,pz ^*^^ 

It has just been shown that the counter E. M. F. E^ is 
nearly equal to the E. M. F. impressed on the armature. 
If, therefore, the E. M. F. applied to the armature is low- 
ered while the field is maintained at normal strength, it is 
evident from formula 13 that 6' will decrease. Also, if the 
applied E. M. F. is increased, the speed will increase. This 
affords one method by which the speed of a shunt motor 
may be varied, namely, by varying the E. M. F. applied to 
the brushes. 



Digitized by VjOOQIC 



22 



DIRECT-CURRENT MOTORS 



§16 



31« In formula 13 we may keep the applied E. M. F. 
approximately constant and vary the speed by changing the 
value of ^. The larger the value of ^, the lower will be 
the speed. In other words, strengthening the field reduces 
the speed, because with a strong field the armature does 
not need to revolve as fast in order to generate the required 
counter E. M. F. Conversely, weakening the field, decreas- 
ing the value of ^, causes the motor to soeed up, because 
the armature has to revolve faster in the weak field to gen- 
erate the counter E. M. F. The quantities ;«, /, and Z in 
formula 13 are fixed for a given motor so that, in general, 
we are restricted to the two means of varying the applied 
E. M. F. and varying the field strength in order to secure 
variations in speed., 

32. Rheostatic Control. — The most common method 
of regulating the speed of a shunt motor is to insert an 
adjustable resistance R^ Fig. 10, in series with the arma- 
ture. This resistance cuts down the applied voltage while 

: jfo<^ . 




ysisom^msi^i^ 



PIO. 10 



the field excitation remains approximately constant. The 
voltage impressed on the armature is less than the line 
voltage by the amount of the drop in the rheostat, and 
it is evident that for a given position of the rheostat, 
this drop, and hence the speed, will depend on the load 



Digitized by VjOOQ IC 



§15 DIRECT-CURRENT MOTORS 23 

on the motor. If the motor is running light, very 
little current will be required to keep it going, and the 
rheostat, even if it is all in, will cut down the speed but 
little. On the other hand, if the motor is heavily loaded, 
the current will be large, and a comparatively small amount 
of resistance in the rheostat will produce a considerable 
change in speed. When a rheostat is in series, therefore, 
every change in load will be accompanied by a change in 
speed, and this is a decided objection to the rheostat control 
for some kinds of work. Moreover, the product of the cur- 
rent by the volts drop in the rheostat represents so much 
waste power, and the rheostat method of control is not an 
economical one; nevertheless, it is very extensively used 
because it is simple and readily applied. Where a large 
number of motors are to be fitted for speed control, it is 
better to use some such arrangement as the multivoltage 
system described later on. Where a single motor is to be 
controlled, it is often cheaper to install a rheostat than 
a more elaborate arrangement, even if the rheostat is some- 
what wasteful of power. When large motors are to be con- 
trolled, the wastefulness of the rheostatic method becomes 
of more concern. 

33. Field Control. — If the speed control is accomplished 
by varying the field strength, a rheostat is inserted in the 
field circuit, the armature being supplied directly from the 
mains. This method is much more economical than where 
the rheostat is in the armature circuit, because with a shunt 
motor the field current is very small and the loss in the field 
rheostat is also small. Unfortunately, however, only a 
limited range of speed variation can be obtained by this 
method, because the fields cannot be strengthened very 
much on account of magnetic saturation, and if they are 
weakened below a certain limit there is sure to be sparking 
at the commutator. Field regulation is therefore limited 
to those cases where a comparatively small range of speed 
variation is required; usually a variation of about 25 per 
cent, can be obtained with an ordinary motor. Motors 



Digitized by VjOOQ IC 



24 DIRECT-CURRENT MOTORS §16 

designed especially for use with the field method of con- 
trol may admit a range of as much as 40 to 50 per cent., 
but such motors must have large fields and are more expen- 
sive than ordinary motors of the same output. The sub- 
ject of speed regulation will be taken up further after the 
other classes of motors have been considered. 



SERLBS MOTORS 

34. These motors are constructed in the same way as 
series-wound dynamos. The most extensive use of series 
motors is in connection with street railways. They are also 
used largely for operating hoists, cranes, and other machin- 
ery of this class that requires a variable speed and strong 
starting effort. Nearly all series motors, like shunt motors, 
are operated on constant-potential circuits. 



SERIES MOTOR ON CONSTANT-POTENTIAl. CIRCUIT 

36. Let A, Fig. 11, represent the armature of a series 
motor connected in series with the field /^across the mains, 




F 



Pio. 11 



as shown. The pressure between the mains is maintained 
constant. We will denote this constant line pressure by E. 
We must have, then, the following relation: 

E=E^^IR,-\-IRf (14) 



Digitized by VjOOQIC 



r 



§15 DIRECT-CURRENT MOTORS 26 

where E^ is the counter E. M. F. of the motor, / the cur- 
rent corresponding to any given load, and R^ and Rf the 
resistances of the armature and field, respectively. 

36. First, we will consider the case where the motor is 
running light. Under this condition of load, the motor will 
take just enough energy from the line to make up for the 
losses due to friction, core losses, etc. As the armature 
speeds up, the counter E. M. F. increases and the current 
rapidly decreases. Now the field is in series with the arma-. 
ture, so that as the current decreases, the field strength 
also decreases, and the armature has to run still faster to 
generate its counter E. M. F., which at no load is just about 
equal to the E. M. F. between the mains. The current 
necessary to supply the losses is usually very small if the 
motor is well designed; consequently, the no-load current is 
very small, and the speed necessary to generate the counter 
E. M F. becomes excessively high. In many cases this 
speed might be high enough to burst the armature. On 
account of this tendency to race, it is not safe to throw the 
load completely off a series motor unless there is some safety 
device for automatically cutting off the current. Of course, 
in street-railway work, or in the operation of cranes, 
hoists, etc., there is always some load on the motors, so that 
no injury from racing is liable to result. 

37. When the motor is loaded, the counter E. M. F. 
decreases slightly, and this allows more current to flow. 
This current strengthens the field, and a correspondingly 
strong torque is produced. It should be noted here that 
the torque of a series motor depends directly on the current 
flowing through it. This quality renders the series motor 
valuable for street-railway work, as a strong starting torque 
can be produced by allowing a heavy current to flow 
through the motor while the car is being started. The 
series motor is capable of exerting a much more powerful 
starting torque than the shunt motor, because of the heavy 
current that flows through the field. In the case of the 



Digitized by VjOOQ IC 



86 DIRECT^CURRENT MOTORS §18 

shunt motor the field current is limited to the normal 
amount by the resistance of the coils, and, consequently, 
this strong magnetizing effect at starting cannot be 
obtained as in the series motor. Since the field strength of 
a series motor increases as the load is applied, it follows 
that the speed will decrease with the load and there will be 
a different speed for each load. This variable speed renders 
the series motor generally unsuitable for stationary work, 
such as operating machinery, etc., but is an advantage for 
street-railway work, where a wide range of speed is desired. 
These advantages regarding starting torque and variable 
speed also apply to cranes, hoists, and some kinds of rolling- 
mill machinery. Series motors are more substantial and 
slightly cheaper to build than shunt motors, on account of 
the fine field winding required by the latter. The field 
coils of series-motors consist of a comparatively small 
number of turns of heavy wire, making a coil that is less 
liable to burn-outs than the fine-wire shunt coils. In short, 
then, the series motor is well adapted for those classes of 
work requiring a large starting torque and variable speed, 
or where variable speed is at least not an objection. 

38. Speed and Torque Curves. — The curves, Fig. 12, 
show the general relation between the torque and speed of a 
series motor. The motor is run from constant-potential 
mains and the load varied. At no load, the current taken 
by the motor is represented by Oa^ and this current multi- 
plied by the voltage represents the power necessary to over- 
come the frictiotial and core losses. Since the current is 
small, the magnetization is low and the speed high, the 
speed being represented by the ordinate O b. With the 
current Oa^ the torque is Oc. As the motor is loaded, 
the torque has to increase, and in order to supply the 
increased torque, the motor has to take more current from 
the line. As the torque and current increase, the speed 
falls at first very rapidly, and then more slowly as the 
fields become saturated, as indicated by the speed curve. 
It will be noticed that the torque increases in almost direct 



Digitized by VjOOQ IC 



§16 



DIRECT-CURRENT MOTORS 



27 



proportion to the current, and that the speed undergoes 
wide variations with changes in the load. 



IT 




»-4: 




t 




r 




I 




^ 




•• t ■ 




1 5 




* V 


2 


1 ^^ 


1 / 


1 si 


^ 


J T^ y 


r^ 


1 ^>-^ 




1 -^^^ 


^^ 


-.Z 




7^t 




-.^f- 




7 




7 




y 




e ^ 


•A« 



Pig. 19 

39. Series Motors In Series Across Constant-Poten- 
tial Mains. — Sometimes the attempt is made to operate 
two or more series motors in series across constant-poten- 
tial mains, as indicated in Fig. 13. For example, 110-volt 
series-wound fan motors are sometimes connected two in 
series across 220 volts. Under such conditions the opera- 
tion of the series motors is unstable. One of the motors 



imss 



Tmxr 



PlO. 18 



may run above its usual speed, and the other run corre- 
spondingly lower, or one may even stop and the other run 
at a speed very much above normal. Moreover, any change 
in the load of either motor will cause a change in the opera- 
tion of both. If the load on each motor were absolutely 



Digitized by VjOOQ IC 



28 



DIRECT-CURRENT MOTORS 



§16 



equal, they would run at the same speed and each would 
utilize half of the power supplied from the mains. If, 
however, one motor is slowed up a little due to a slight 
increase in load, the other will speed up, thus increasing the 
E. M. F. across its terminals and depriving the other motor 
of its E. M. F. The result is that the first motor is greatly 
slowed down or may even stop. This unstable operation is 
done away with if the two motors are geared together, so 
that they are compelled to run at the same speed. On 
street cars, series motors are operated at slow speeds in 
series, but they are both practically geared together because 
they are both geared to the axles. Their operation is there- 
fore stable unless the wheels slip on the rails. If the same 
motors were set up on the floor and not geared together in 
any way, they would show the unstable behavior referred 
to above. 



SPSEB BEGUULTIOK OF SEKEES MOTORS 

40, In some cases it is necessary to have the speed of a 
series motor under control, and this is nearly always accom- 
plished by a rheostat in series with the motor, as shown in 

. . JCsift 



r(mmm)^ 



L^l 



jCiaEliL. 




PlO. 14 



Fig. 14. Sometimes the field coils are wound in sections 
and a switch arranged so that these sections can be cut in 



Digitized by VjOOQIC 



§16 DIRECT-CURRENT MOTORS 29 

or out, thus varying the speed through a limited range by 
changing the field strength. In other cases a variable 
shunt is connected across the field coils. On street cars the 
speed is varied by changing the voltage applied to the 
motors. The ordinary street car is equipped with two motors 
designed to operate normally on 500 volts. For low speeds 
these motors are connected in series across the 500 volts, 
thereby making the voltage applied to each motor approxi- 
mately 250 volts and giving a greatly reduced speed. 
When a high speed is desired, the connections are changed 
by means of the controller so that the motors are connected 
in parallel across the line. This gives the full voltage 
across each of the motors, and allows them to run at high 
speed. This method of series-parallel control avoids the 
use of resistance for low speeds, and is therefore more 
economical of power than the rheostatic method; it will be 
described more in detail in Electric Railways. 

41, The field magnets of series motors are sometimes 
provided with such a large number of turns that the field 
becomes fully magnetized when the current flowing is only 
a fraction of the full-load current of the motor. This gives 
a field that does not change greatly in strength for a con- 
siderable range of load, and thus tends to make the speed 
vary less with changes in the load, and also keeps the motor 
from sparking at moderate loads, on account of the strong 
field obtained. Street-railway motors are generally over* 
wound in this way. 



SERIES MOTOR ON CONSTANT-CURRENT CIRCUIT 

42, It has already been mentioned that series motors 
have been operated to a limited extent in the past on con- 
stant-current circuits. Fig. 15 shows a series motor M con- 
nected in an arc-light circuit, the current in which is kept at 
a constant value by means of a regulator on the dynamo D, 
'Since the current flowing through the field F of the 
motor is constant, the strength of field will be constant 



Digitized by VjOOQ IC 



30 DIRECT-CURRENT MOTORS §16 

and the torque will also be constant. If the arma- 
ture were allowed to run free, it would race very badly, 
the racing being worse than in the case of a series motor on 
a constant-potential circuit, because in the latter case the 
current in the field and armature is reduced and the torque 
correspondingly cut down as the speed increases, whereas 
in this case the current is kept constant and the torque 
remains the same. It is thus seen that the speed of a series- 
motor operated in this way would vary widely with the load, 




Pig. 16 

and such a motor without some regulating device would be 
unsuitable for operating machinery. Motors are now sel- 
dom if ever operated on arc circuits. Series motors on arc 
circuits are always more or less dangerous on account of the 
high pressures generated by arc dynamos. Also, if the 
output of the motor is at all considerable, the pressure 
across its terminals at full load must be very high because 
the current is small. 



COMPOUISD-WOUND MOTORS 



DIFFBRENTIAL.t.Y WOtTND MOTORS 

43. These motors are essentially the same in construc- 
tion as the compound-wound dynamo, except that the 
series-coils are connected so as to oppose the shunt coils 
instead of aid them as in the dynamo. The object of 
this arrangement is to secure constant speed when the 



Digitized by VjOOQ IC 



§15 



DIRECT-CURRENT MOTORS 



31 



voltage of the dynamo supplying the motor is constant. 

These motors are not ^ ^ 

used to any great ex- Maim 

tent, because it is found 
that a well-designed 
shunt-wound motor will 
give sufficiently close 
speed regulation for 
all practical purposes. 
Fig. 16 shows the con- 
nections of a differen- 
tial motor, the coils 
being intended to rep- 
resent windings in pio.w 

opposite directions, one right hand, the other left hand. 




ACCrrMITLATIVBLY WOUND MOTORS 

44, The accumulatively wound motor is the same in 
construction as a compound-wound dynamo. The differ- 
ence between an accumulatively wound motor and a differ- 
entially wound one lies in the way in which the series-coils 
are used. In the accumulatively wound machine the series- 
coils aid the shunt coils in magnetizing the field, whereas in 
the differentially wound motor the reverse is the case. The 
accumulatively wound motor combines the features of the 
shunt motors and series motors. The series-coils are added 
in order to obtain a more powerful starting torque than that 
given by a plain shunt motor. These series-coils are usu- 
ally cut out after the machine has been started, and the 
motor then runs with the shunt coils alone. The motor 
thus combines the strong starting torque of the series 
motor with the valuable constant-speed qualities of the 
shunt motor. Accumulatively wound motors are partic- 
ularly adapted for the operation of printing presses, eleccric 
elevators, or other machinery requiring a strong starting 

44—19 



Digitized by VjOOQ IC 



32 DIRECT-CURRENT MOTORS §16 

effort, but where the variable speed of the series motor would 
be objectionable. The conditions to be met in the opera- 
tion of printing presses and similar machinery are particu- 
larly hard to fill, because a very strong starting effort is 
needed on account of the large amount of friction caused 
by gears, rollers, etc. After the motor has been started, a 
much smaller torque is sufficient. Speed control is usually 
obtained by inserting resistance in series with the armature 
for low speeds, and by cutting out the resistance for higher 
speeds. The series-coils are frequently wound in sections 
so that they can be cut out step by step, thus further rais- 
ing the speed by weakening the field. Compound-wound 
motors are thus desirable for certain lines of work, but they 
are more expensive and complicated than shunt motors. 
For the majority of stationary machines the shunt motor is 
well suited, and on account of its simple construction and 
consequent cheapness is used much more largely than com- 
pound-wound motors, which are more in the nature of 
special machines. 



BYNAMO AND MOTOR ROTATION 

45, Very often it is desired to convert a dynamo into a 
motor, and the question arises as to what changes in con- 
nections are necessary. The changes required will depend 
on the direction in which the motor is required to run. 
Usually the machines so converted are shunt-wound, though 
sometime^ they are compound-wound. 

46, Shunt Dynamo Converted to Shunt Motor. — 

Fig. 17 {a) shows a machine supposed to be operating as a 
shunt dynamo. The armature is wound so that when the 
machine is driven in the direction indicated by the arrow 
and the polarity of the fields is as shown, the left-hand 
brush is positive, and the current flows out on line i. Con- 
sider a conductor a on the left-hand side of the armature, as 
shown in the small lower figure of (a). With the direction of 



Digitized by VjOOQ IC 



§1* 



DIRECT-CURRENT MOTORS 



rotation as indicated, this conductor will be moved up across 
the field as shown by the vertical arrow, and the current 
set up in a will therefore be directed downwards. The wire 
in its motion may be looked on as bending the lines of force, 
and thus tending to set up right-handed magnetic whirls 
around itself, as indicated. Now suppose that the same 
machine, without any change whatever in the connections, 
is supplied with current from an outside source, as shown 
in Fig. 17 (b). We will also suppose that line 1 is connected 
to the positive main, so that the current now flows in at the 
right-hand brush. The current in the armature is there- 
fore the opposite to that in (a), but the field remains excited 




PIO. 17 

in the same direction as before. The current in conductors 
now flows up and sets up left-handed whirls, as shown by 
the dotted circles. These strengthen the original field 
below the conductor and weaken it above, thus forcing 
the wire up. In other words, the direction of rotation 
is the same as when the machine was driven as a dynamo. 
This property of the shunt machines is of importance 
when machines are run in parallel, because if one 
machine is motored by current sent back through it from 
another, it simply runs as a motor in the same direction 
in which it was driven as a dynamo and no disturbance 
results. 



Digitized by VjOOQ IC 



84 



DIRECT-CURRENT MOTORS 



§16 



47. If the lines 1 and 2 are reversed in their connection 
to the positive and negative mains, as shown in Fig. 17 (r), 
the direction of rotation still remains the same because the 
current is reversed in both armS:S;e and field. If the motor 
is required to run in a direction opposite to that in which it 
operated as a dynamo, the field terminals should be reversed. 
To change a shunt dynamo to a shunt motor, then, no 
change in connections is necessary if the direction of rota- 
tion is to remain the same, though, of course, it is necessary 
to insert the necessary starting box, as described later. 

48, Series Dynamo Converted to Series Motor. — 

Fig. 18 {a) shows a machine operating as a series dynamo 
under conditions similar to those of Fig. 17 (a). If this 



11 



ssnom) 



t 




't 



*i+ 







w 



FiO. 18 




machine is run as a motor having line 1 connected to the 
positive main, as shown in {b), the direction of rotation will 
be reversed, because the machine is now running as a 
motor, and the field current is also reversed in direction. 
The action is therefore opposite to that of the shunt motor, 
in which the direction of rotation remains unchanged. The 
direction of rotation still remains reversed if the connec- 
tions to the mains are changed, as shown in Fig. 18 {c). If, 
therefore, a series machine were operating as a dynamo, 
and if for any reason current should flow back through it, 
the direction of rotation of the machine would be reversed 
and damage would result. If the direction of rotation as a 



Digitized by VjOOQ IC 



§16 DIRECT-CURRENT MOTORS 85 

motor is to be kept the same as that of the dynamo, either 
the field or armature connections must be reversed. 

49, Compound Dynamo Converted to Motor. — A 

compound dynamo can be converted into a motor of 
either the differential or accumulatively wound type. In 
many cases compound dynamos are converted to motors, 
the series-coils are cut out altogether, and the machine 
operated as a plain shunt motor. Since in a compound 
dynamo the series-coils aid the shunt, it follows that if the 
machine is operated as a motor without any change what- 
ever in connections, the series-coils will oppose the shunt, 
and a differential motor will be obtained; if the series- 
coils are sufficiently powerful to overpower the shunt, the 
direction of rotation will be reversed. As differentially 
wound motors are seldom required, it will usually be neces- 
sary to reverse the series-coils connections with respect to 
the shunt when the machine is used as a motor, provided 
the series-coils are used at all. 



AUXILIARY APPARATUS 



STARTrN'G BHEOSTAT8 

60, When motors are operated on constant-potential 
circuits, it is necessary to insert a resistance in series with 
the armature when starting the motor. Of course, in the 
case of a series motor, this starting resistance is also in 
series with the field. The resistance of a motor armature is 
very small, and in the case of a series motor the field resist- 
ance is also small, so that if the machine were connected 
directly across the circuit while standing still, there would 
be an enormous rush of current, because the motor is gen- 
erating no counter E. M. F. Take, for example, a shunt 
motor of which the armature resistance is .1 ohm. If this 
armature were connected across a 110-volt circuit while the 



Digitized by VjOOQIC 



36 DIRECT-CURRENT MOTORS §15 

motor was at a standstill, the current that would flow 

momentarily would be -— = 1,100 amperes, the amount 

being limited only by the resistance of the armature. In 
the case of a series motor, the rush of current would not be 
quite as bad, as the field winding would help to choke the 
current back. 

51. The starting: rheostat, or starting: box, as it is 

often called, is a resistance divided up into a number of 

sections and connected to a switch 
by means of which these sections 
can be cut out as the motor comes 
up to speed. When the motor is 
running at full speed, this resist- 
ance is completely cut out, so that 
no energy is lost in it. Fig. 19 
shows a simple form of motor- 
starting rheostat, the resistance 
wire in this particular type being 
bedded in enamel on the back of 
Pig. i» an iron plate, while the ribs r on 

the front are intended to present additional cooling surface 
to the air. Starting rheostats are not designed to carry 
current continuously, and should therefore never be used 
for regulating the speed of the motor. The resistance wire 
is made of such a size as to be capable of carrying the cur- 
rent for a short time only, usually 15 to 30 seconds, and if 
the current is left on continuously, the rheostat will be 
burned out. The handle k of the rheostat shown is pro- 
vided with a spiral spring s^ tending to hold it against the 
stop ^, which makes it impossible to leave the contact arm 
on any of the intermediate points. On the last point a 
clip c is placed to hold the arm of the rheostat. Starting 
rheostats are made in a great variety of forms and sizes, 
but the object is the same in all of them, i. e., to provide a 
resistance that may be inserted when the motor is at rest 
and gradually cut out as the motor comes up to speed. 



Digitized by VjOOQ IC 



J 15 



DtkECT-CURRENT MOTORS 



37 



SHXmr-MOTOB CONNECTIONS 

62. One method of connecting a shunt motor to con- 
•stant-potential mains is shown in Fig. 20. The lines lead- 
ing to the motor are 
connected to the mains 
through a fuse block Z>, 
from which they are 
led to a double-pole 
knife switch B, A 
circuit-breaker is to 
be preferred to the 
fuse block, especially 
if the motor is a fairly 
large one. One end 
of the shunt field F is 
connected to ter- 
minal 1 of the motor, 
and one brush is also 
connected to the same 
terminal. The other 
field terminal is con- 
nected to the motor 
terminal ^, and the 
other brush leads to 
the third terminal S, ^o- » 

One side of the main switch connects to terminal 1\ the 
other side connects to S through the starting rheostat C. 
Terminal 2 connects to the same side of the switch as the 
starting rheostat. It will be seen from the figure that as 
soon as the main switch is closed, current will flow through 
the field F. When the rheostat arm is moved over, current 
flows through the armature, and the* motor starts up. The 
handle is then moved over slowly and left on the last point 
when the motor has attained its full speed. 

53. In Fig. 20 one terminal of the field is connected to 
one of the brushes, and three terminals only are provided 




Digitized by VjOOQ IC 



38 



DIRECT-CURRENT MOTORS 



§16 



on the motor. A motor connected in this way will always 
run in a certain direction no matter which line wires are con- 
nected to terminals 1 and S. Motors are now, almost with- 
out exception, provided with radial carbon brushes, and 
they can be run in either direction. It has become common 
practice, therefore, to provide them with separate field 



Sif/^ Af0//fs 



Orcurf^ Brea/ter 
or fi/s0s — 



Cfrcw'/ Breaker 
orfUses — 





PIO. 21 



FIO. S8 



and armature terminals, as shown in Fig. 21. By connect- 
ing field terminal 2 to armature terminal 1 by means of 
wire ^, the motor will run, say, right-handed, as shown by 
the arrow. If it is desired to connect the motor for the 
reverse direction of rotation, it is easily done by joining 
terminals 1 and S and connecting field terminal 2 to the 
switch. This reverses the current in the field, as shown in 
Fig. 22. 

54. Blieostats With Automatic Release. — In the pre- 
ceding diagrams the simplest type of rheostat has been 
shown in order to make the connections as clear as pos- 
sible. The simple starting rheostat shown in Figs. 19 and 
20 is now used but little, for the following reasons. Sup- 
pose a motor is running and that the attendant shuts it 



Digitized by VjOOQ IC 



§15 



DIRECT-CURRENT MOTORS 



39 



move 



down by opening the main switch, but forgets to 
the rheostat arm back to the 
off-position. When the motor 
is started again, the chances 
are that it will not be noticed 
that the rheostat is at the on- 
position, and when the main 
switch is closed a rush of 
current that may damage the 
motor takes place. Again, the 
motor may be running and 
the power may be thrown off 
the line for some reason or 
other, and when it is thrown _ _ 

PlO. 28 

on again, the rheostat is at the 

on-position and a rush of current takes place. For these 





Pig. M Ar/7»a/bt€ 

reasons it is customary to install automatic rheostats that 



Digitized by VjOOQIC 



40 DIRECT-CURRENT MOtORS § lU 

will fly back to the off-position whenever the power is cut 
off from the motor, either by opening the main switch or by 
the power going off the line. Fig. 23 shows a simple form 
of automatic rheostat made by the General Electric Com- 
pany. The automatic feature consists of an electromag- 
net A in series with the motor field. The lever C is moved 
over against the action of a coiled spring, and is held at the 
on-position by the attraction of magnet A for the afmature 
B, Fig. 24 shows the rheostat connected to a motor. If the 
current supply is interrupted, the current in coil A gradually 
decreases as the motor slows up, and its pull becomes 
weaker, until finally the armature B is released, and the arm 
flies back to the off-position. The coil A is connected in the 
field circuit rather than in series with the armature, because 
it protects the motor in case the field circuit becomes broken. 
If the field circuit were broken and the armature left con- 
nected to the mains, a large rush of current would take place 
because the breaking of the field circuit wpuld reduce the 
counter E. M. F. to practically zero. 

55. Methods of Ck>niiectiii^ Shunt CJoll. — Practice 
differs in some respects regarding the connections for shunt 
motors. The difference lies in the method of connecting 
the shunt field. If the student will examine Figs. 20 
and 26, he will notice that there is a difference in the con- 
nections of the shunt winding. The connections of Fig. 20 
are equivalent to those in Fig. 26 (^), while those of 
Fig. 26, are shown in (*), Fig. 25. In (a) the shunt field is 
connected so that as soon as the main switch is closed, cur- 
rent flows through the shunt field. The field is therefore 
excited before current is allowed to flow through the arma- 
ture by moving the rheostat to the first point. In {p) the 
field is not excited until the rheostat arm is placed on the 
first contact point. As the arm is moved over, the field 
current has to back up through the resistance, and when 
the arm is at the on-position, the starting resistance is in 
series with the field. On many boxes where this scheme of 
connections is used, an auxiliary contact a is arranged at 



Digitized by VjOOQ IC 



§15 



DIRECT-CURRENT MOTORS 



41 



the end of the travel of the arm, so that the shunt current 
will not pass through the starting resistance when the 
motor is in regular operation; but even if this contact is 
not provided, the starting resistance is usually so low and 
the field current so small that its insertion in series has 
little appreciable effect on the operation of the motor. The 
chief advantage of the connections shown at {b) is that they 



I 

I 



f-^OOOOOOOO'^ 



ffdease Magnet 



•Q--J O^AAAAAAArl pL-"?- 



mooms 




I 



\ffefea5€Ma^n&t 



^Qjr^ (^lAAAAAAArl j 




avoid breaking the field circuit of the motor. As soon as 
the rheostat arm, Fig. 26 (^), flies back, the field circuit is 
opened. This always causes more or less of an arc and 
consequent burning action on the last rheostat contact, but 
a more objectionable feature is that the sudden stoppage of 



Digitized by VjOOQ IC 



42 



DIRECT-CURRENT MOTORS 



§16 



the shunt current is liable to set up such a high induced 
E. M. F. in the windings that there is danger either of 
weakening or breaking down the field insulation. With the 
connection shown at (b) the field circuit is never broken; 
when the arm flies back there is scarcely any percepti- 
ble sparking and there is no strain on the field winding. 
The principal argument advanced in favor of the connec- 
tions shown at (a) is that the field is fully excited before 
current is allowed to flow through the armature, and as it 




PlO. 96 



takes the current in the shunt winding a short time to 
build up, a better starting torque is secured than when the 
connections (6) are used. This may be the case with large 
motors, but for motors of ordinary size, the time required 
for building up the field is so small as to be of little impor- 
tance in this connection. Fig. 26 shows a Cutler-Hammer 
rheostat connected according to plan (d), which is the one 
used generally by the Cutler-Hammer Company on their 
shunt-motor rheostats. 



Digitized by VjOOQ IC 



§15 



DIRECT-CURRENT MOTORS 



48 



56. Wrong: Ck>iiiiectioii of Shunt Field. — Perhaps the 
most common mistake made when connecting up shunt- 
motor starting boxes is that shown in Fig. 27. This is the 
outfit for which the correct connections are shown in Fig. 26. 
The mistake consists in interchanging the heavy wires at 
the starting box, the line wire being placed in the armature 
binding post and the armature wire in the line binding post. 




FlO. 2? 



Pig. 25 (c) shows these incorrect connections in a simplified 
form. As soon as the main switch is closed and the rheo- 
stat placed on the first point, the field is connected directly 
across the brushes, and the pressure applied to the field is 
therefore equal to the difference of potential between the 
brushes. Now, at starting, nearly the whole of the applied 
pressure is taken up in the starting resistance, and the only 
pressure across the field terminals is that due to the drop in the 
armature. The result is that the field is excited to but a 
very slight extent and the motor refuses to start. If the 



Digitized by LjOOQ IC 



44 DIRECT-CURRENT MOTORS §16 

rheostat arm is moved over, the current increases, and if it 
is moved far enough, may become sufficiently large to burn 
out the box. In some cases, if the load is not heavy, the 
motor may start up after the arm has been moved over a 
considerable distance, because the current may then become 
so great that the pressure across the brushes will excite the 
field enough to give the required torque. If, after con- 
necting up a starting box, it is found that the motor does 
not start promptly on the first or second point, the con- 
nections should be carefully traced out to see that the 
above mistake has not been made. It is a very easy matter 
to get these wires confused, especially if they are run 
through conduit or bunched together in any way. This 
mistake is frequently made when the motor is situated some 
distance from the starting box, as the wires are then easily 
mixed. 

67. Types of Automatic-Release Starting Rheo- 
stats. — Fig. 28 shows one of the smaller types of the Cutler- 
Hammer rheostat with automatic release. The connections 



Pig. S8 Pio. 99 

for this rheostat are shown in Fig. 26. Fig. 29 shows a 
General Electric starting rheostat with automatic release, 
and also provided with an overload attachment that causes 



Digitized by VjOOQ IC 



SIS 



DIRECT-CURRENT MOTORS 



45 



the arm to fly back to the off-position in case the current 
becomes excessive. Fig. 30 shows the connections for this 
box; it will be noted that it is connected according to 
Fig. 25 {a), the field being excited as soon, as the main 
switch is closed. In Fig. 30 the rheostat arm is in the 
running position, all the resistance being cut out. By 
arranging the resistance as shown, the current flows across 
the contact arm and does not have to pass through the 
pivot. The contact arm is moved over against the action 




Ksmmmmj 

Pio. 80 

of a spiral spring in the hub and is held in position by a 
catch a that fits into a notch in the hub of the lever d. 
This lever carries an armature r, which is held down by 
the release magnet m. 

The device for protecting the mojtor against overloads 
consists of an electromagnet, the coil of which is connected 
in series with the armature A, This magnet is provided 
with a movable armature ^, the distance of which from 



Digitized by VjOOQ IC 



46 DIRECT-CURRENT MOTORS §16 

the pole h may be adjusted by the screw k. When the 
current exceeds the allowable amount, the armature is 
lifted, thus making connection between the pins /. This 
connection short-circuits the coil of the magnet m and the 
lever goes to the off-position. 

68. The design of a starting rheostat depends very 
largely on the size of the motor with which it is to be used, 
since heavy currents require large contact surfaces and 
heavier construction throughout. Speed-regulating rheo- 
stats are constructed in much the same manner as starting 
rheostats, the only difference being that the regulating 
rheostat is much larger, so as to stand continuous use 
without overheating. 

REVERSING DIRECTION OF ROTATION 

69. It has already been mentioned that if the current 
in either the field or armature of a motor be reversed, the 
direction of rotation will be reversed. This is evident from 
an inspection of Figs. 2 and 3. If the current in wire a be 
reversed while the field is left unchanged, the direction of 
motion will be reversed. Also, if the direction of the cur- 
rent in the wire is left unchanged, and the field reversed, 
the direction of motion will be reversed. If both current 
and field be changed, the direction of motion will remain 
unchanged. 

60. A series motor will run in the same direction, no 
matter which of the supply lines is connected to its ter- 
minals tf, d. Fig. 31. Reversing 
the line connections simply re- 
verses the current through both 
armature and field, and does not 
therefore change the direction 
of rotation. In order to reverse 
the motor, either the armature 
^'®' ** terminals r, d must be inter- 

changed, so as to reverse the current through the armature, 
or the terminals d^ b must be interchanged, so as to reverse 




Digitized by VjOOQ IC 



§15 



DIRECT-CURRENT MOTORS 



47 



the current through the field. In street-railway work the 
motors are usually reversed by reversing the current 
through the armature, the current through the field remain- 
ing unaltered. 

When it is desired to reverse a motor while it is running, 
it is very necessary to insert a resistance in the armature cir- 
cuit before reversing the current through the armature. It 
must be remembered 
that the counter 
E. M. F. which the 
motor was genera- 
ting just before 
reversal becomes an 
active E. M. F. and 
helps to make the 
current flow through 
the armature as soon 
as the current is re- 
versed, and this 
action continues until 
the motor starts to 
turn in the opposite 
direction. It is best, 
therefore, when pos- 
sible, to let the motor 
drop considerably in 
speed, or even come 
to a standstill, before 
reversing it. 

61. ReversinfiT 

Sipvltclies. — In order 

to reverse the current 

in a motor armature 

so as to bring about 

a reversal of rotation, 

a reversing switch is placed in the armature circuit. 

Fig. 32 shows three common types of switch. That shown 

44—20 




Digitized by VjOOQIC 



48 



DIRECT-CURRENT MOTORS 



§16 



at (a) consists of two metal blades a, b hinged at r^/ and 
connected together by an insulating cross-piece e. The 
blades can be moved from the position shown in the figure 
to that indicated by the dotted lines, by pujling on the 
rod f. In one position, c and d are connected to g and A, 
while in the other position they connect to h and >fe, thus 
reversing the current in the armature. Fig. 32 {b) shows 
an ordinary double-pole double-throw knife switch used as 
a reversing switch. The middle clips are connected to 
the armature, and the top and bottom clips are cross-con- 
nected, so that when the 
switch is thrown up, the 
current in the armature 
is in one direction, and 
when it is thrown down, 
the armature current is 
reversed. The revers- 
ing switch shown at (c) 
is of the cylinder type, 
and is used very 
largely for street-car 
controllers. The upper 
sketch shows the cylin- 
der and indicates the 
arrangement of the con- 
tacts. Contact fingers 
are connected to termi- 
nals a, b, c, d^ as shown 
by the end view in the 
lower sketch. Contact 
pieces e^f, g, A, ^, / are 
mounted on a wooden 
drum that can be ro- 
tated through an angle 
sufficient to bring the 
fingers in contact with 
either set of plates. When the fingers rest on plates e and /, 
as indicated by the dotted line i, the current entering at a 




Pig. W 



Digitized by VjOOQ IC 



§15 



DIRECT-CURRENT MOTORS 



49 



takes the path a-^e-b through armature c-f-d. When the 
drum is turned so that the fingers rest on the other contacts, 
as indicated by the dotted line ^, the path becomes a-g-k-c 
through armature in the reverse direction b-h-l-d. 

62. Shunt Motor Witli Reversing Swlteli. — Fig. 33 
shows connections for a shunt motor with reversing switch R, 
The field is excited from the mains as soon as the rheostat 
is placed on the first point, and remains excited in the same 
direction regardless of the position of the reversing switch. 

63. Fig. 34 shows a Cutler-Hammer le verse starting 
rheostat. When the motor is not running, the handle h is 




PIO.84 



held at the central position and the motor can be started in 
either direction by moving the handle to one side or the 



Digitized by VjOOQIC 



50 



DIRECT-CURRENT MOTORS 



§16 



other, thereby changing the direction of the current in the 
armature circuit. The arm is held at the on-position by 
the catch on the end of the armature of the release magnet r 
engaging with the notches n or b. This starter is provided 
with auxiliary laminated contacts r, rf, e^ f^ which press 
against the studs g^ g^ thus giving a good contact for carry- 
ing the current continuously aft«r the motor has been started. 
Shunt motors are always reversed by reversing the current 
in the armature, because it is not advisable to interrupt 
or reverse the shunt field current on account of the high 
induced E. M. F.'s that are liable to be set up. 



SERIES-MOTOR COXNECTIOKS 

64. The connections for series motors are on the whole 
simpler than those for shunt motors. Since the field is 



' \Cuvuff 



i Cutout 



i^fifzk 




'W33S33T 

PlO.85 



r^r^ 



Armafurm 



in series with the armature, and helps to choke back the 
current at starting, a series motor does not require as large 



Digitized by VjOOQIC 



§15 



DIRECT-CURRENT MOTORS 



61 



a starting resistance as the shunt motor. Fig. 35 shows a 
simple series-starting rheostat and connections. 

For motors that have to be stopped, started, and reversed 
frequently, special types of starting devices are used. These 
are generally called controllers. For street-railway work 
these controllers become quite complicated, as they are 
designed not only to cut resistance in or out, but also to 
make various combinations of the two or more motors used 
on a car. A full ex- 
planation of these con- 
trollers will be found 
in Electric Railways^ 
so they will not be c<5n- 
sidered here. Control- 
lers somewhat similar 
to those used on street 
cars are also used for 
stationary work, but 
when so used they 
are generally required 
to control but one 
motor, and hence are 
designed to simply cut 
resistance in or out 
and not make series 
and parallel combina- 
tions. Fig. 36 shows a 
series motor equipped 
with an ordinary start- 
ing box and reversing 
switch R, 

66. The series- 
motor, as already 
stated, is used very 
largely for operating traveling cranes, hoists, rolling-mill 
machinery, etc. The use of these motors in rolling mills 
and other places calling for heavy service has resulted in the 




PlO.86 



Digitized by VjOOQIC 



62 DIRECT-CURRENT MOTORS §15 

development of a large number of controlling devices spe- 
cially adapted to work of this kind. For such service the 






motor must be capable of being stopped, started, and 
reversed quickly, and the controller must be of simple and 



Digitized by VjOOQIC 



§15 DIRECT-CURRENT MOTORS 63 

substantial construction. Fig. 37 shows three views o£ a 
controller that is used in a large number of steel mills and 
other plants for the operation of cranes, etc. It is known 
as the Dinkey controller, manufactured by the Electric 
Controller and Supply Company, of Cleveland, Ohio. The 
resistance coils and contact switch are mounted together, 



PlO. 88 

fio that there are but four terminals to the controller, and it 
can therefore be easily disconnected and replaced in case of 
trouble. The resistance coils a, Fig. 37 (r), are wound on 
heavy asbestos tubes and mounted as shown. The terminals 
b of the resistance connect to the brass lugs to which the 
drop-forged copper contacts c are screwed, view (^). When 



Digitized by VjOOQ IC 



64 DIRECT-CURRENT MOTORS §18 

the motor is at rest the operating handle d occupies the 
vertical position ; a forward movement of the handle rotates 
the arm e^ view (^), and causes the motor to run in one 
direction; a reverse movement of the lever reverses the 
rotation of arm e and makes the motor run in the oppo- 
site direction. The contact switch, therefore, serves both 
to gradually cut out the resistance and also to reverse 
the motor. The arm carries four sets of contacts /, g^ A, 
and >fe, insulated from it ; the inner contacts g and h 
bear on the contact arcs /, //f, which are provided with renew- 
able tips o^p. When the 
controller is at the oflf- 
position, the contacts 
rest on the insulating 
pieces r, r, r, r. Contacts 
/ and g^ and h and k are 
connected together by the 
coils s^ J, so that the cur- 
rent passes through these 
coils and sets up a mag- 
netic field in the region of 
the contacts. The object 
of this field is to suppress 
sparking at the contacts; 
as soon as an arc is formed, 
it is forced out or blown 
out by the magnetic field. 

66. Fig. 38 shows the 

connections. The upper 

and lower left-hand sets 

of contacts are connected 

to the two groups of 

resistance coils, and the 

f right-hand contacts are 

connected across to the 

^'®- ^ similar contacts on the 

left, as indicated by the horizontal dotted lines. When 



Digitized by VjOOQ IC 



8 16 DIRECT-CURRENT MOTORS 55 

the operating lever is in the full-line position, the current 
flows as shown by the arrowheads, but when thrown over to 
the dotted position, the current through the armature is 
reversed, while that in the field remains the same, thus 
reversing the direction of rotation. This can be readily 
seen by tracing the path of the current when connection is 
made between the contact arcs and segments as indicated 
by the dotted lines a b^ and c d. 

67. Fig. 39 shows another style of electric-crane con- 
troller made by the same company. It is similar in principle 
to the one just described, the only difference being that the 
four sets of contacts are mounted on two vertical parallel 
slate bases, instead of on a single base, as in Fig. 38. 



68. Fig. 40 {a) shows a front view of a smaller type of 
controller made by the Electric Controller and Supply 
Company. It is intended for lighter work than the two 



Digitized by LjOOQ IC 



56 



DIRECT-CURRENT MOTORS 



8i« 



previously described, and will handle motors up to 17^ horse- 
power at 600 volts. Fig. 40 (d) shows a rear view of the 




PlO. 41 



slate face with the resistance coils in place. Fig. 41 shows 
the method of connecting one of these controllers to a 
series motor. 



AUTOMATIC STARTING RHBOSTATS 

69, Sometimes it is necessary to have a starting rheo- 
stat arranged so that it can be controlled from a distant 
point, in which case the box has to cut out the resistance 
automatically. The most common method of accomplishing 
this is to provide the rheostat with a solenoid which, when 
energized, moves the contact arm and cuts out the 
resistance. 

70, Fig. 42 shows a Cutler- Hammer automatic starter. 
In this case it is used to control a motor that operates a 



Digitized by LjOOQ IC 



§15 



DIRECT-CURRENT MOTORS 



67 



pump supplying water to a tank. When the solenoid a is 
energized, it draws up its core and thus moves the arm b 
over the contacts. The motion is controlled by a dashpot c 
filled with oil, so that the arm is drawn slowly over the con- 
tacts. The controlling switch at the distant point is shown 
at d^ and in this case the switch is opened and closed by a 
float. When d closes, the solenoid switch e operates, thus 




PlO. 48 

raising its core and making contact between studs f^g. 
This closes the main circuit. Solenoid a is energized at the 
same time, and the resistance is gradually cut out, thus 
starting the motor. When the core of a reaches the upper 
position, the carbon points at // are separated, thus placing 
a lamp in series with a and preventing its overheating. 
When the core has reached its upper position, a small cur- 
rent is sufficient to hold it there. After e has closed and 



Digitized by VjOOQIC 



58 DIRECT-CURRENT MOTORS §15 

arm b has moved from its lowest position, the carbon points 
at k separate, thus placing a lamp in series with e. Placing 
lamps or other resistance in series with the solenoids not 

only prevents heat- 
ing, but also effects 
a saving in current. 

71, In the case 
of automatic con- 
trollers designed for 
large currents, it 
has been found that 
the sliding-contact 
method gives more 
or less trouble due 
to roughing up of 
the contacts, and 
that the arm some- 
times sticks instead 
of returning to the 
off-position when 
the current is shut 
off. This is espe- 
cially the case with 
controllers for elec- 
tric elevators where 
the starting and 
stopping is very 
frequent. In order 
to do away with 
sliding contacts, a 
type of controller of 
which Fig. 43 is an 
example is now 
P'c- ^ largely used for this 

class of work. It consists of a number of automatic switches, 
each section of the resistance being cut out by an indi- 
vidual switch. Each switch consists of a solenoid which 



Digitized by VjOOQ IC 



§15 DIRECT-CURRENT MOTORS 59 

when energized raises its core and brings the disk d^ 
mounted thereon into contact with the fingers /,/'. In 
this controller the switch C closes the main circuit, 
while A' and B' control the motion of the elevator up or 
down, that is, A' and B' are reversing switches. The 
smaller solenoids control the various sections of the resist- 
ance and, as compound-wound motors are used for elevator 
work, also cut out the series-field when the motor has 
attained full speed. The main fuses are at k k, and the 
resistance is mounted in a separate case behind the con- 
troller. The contact fingers are provided with auxiliary 
carbon tips ;r, ;r, at which the break takes place. The disks 
are free to revolve, so that whatever burning action there 
may be is distributed around the whole disk, and there is 
therefore little danger of sticking. 



MXJIiTIVOIiTAGE SPEED CONTBOIi 

72. As previously pointed out, the speed of a motor may 
be controlled by varying the E. M. F. applied to the arma- 
ture, and the simplest way of doing this is to insert a 
resistance in series with the armature. For some classes of 
work a wide range of speed control is necessary. For 
example, in machine shops and printing plants this is the 
case. One great objection to the rheostatic method of con- 
trol, outside of its wastefulness, is that a change in speed 
occurs with every change in load. Suppose the motor is 
carrying a given load and that the rheostat is set at a point 
where the required speed is obtained. Now, if the load is 
increased, the current will increase, and the drop in the 
rheostat will also increase, thus cutting down the E. M. F. 
at the armature terminals and decreasing the speed. In 
the same way a decrease in load will cause an increase in 
speed. In order to overcome the disadvantages of rheo- 
static control, a number of so-called multivoltagre sys- 
tems have been devised. As a rule these systems are 
intended for those places where a number of motors have to 



Digitized by VjOOQ IC 



60 



DIRECT-CURRENT MOTORS 



§15 



be operated at variable speed, because they are too expen- 
sive to install for single motors except in some special 
cases. This method of control is sometimes known as the 
Leonard system, on account of the patents relating to it 
having been taken out by Mr. H. Ward Leonard. 

73, Fig. 44 shows a simple case of multivoltage con- 
trol. The dynamo A supplies current to the motor £. 
Current is supplied from the constant-potential mains C to 
the fields D and £ of the dynamo and motor. The motor 
field receives a constant excitation, but that of the dynamo 
can be varied by means of a field rheostat F. It is evident 
that by varying the field excitation of the dynamo, the pres- 
sure applied to the motor can be varied through a wide 




PIO. 44 

range, and since the motor field excitation is constant, a 
correspondingly large range in speed can be obtained. 
This method, therefore, allows the voltage applied to B to 
be varied without the insertion of resistance between B 
and A, For each voltage of A there will be a definite 
speed of B, and this speed will remain practically constant 
no matter what load B is called on to carry. 

74. The method shown in Fig. 44 is not generally 
applicable, because it requires a dynamo for each motor, 
and this is out of the question, except, perhaps, in a few 
special cases. However, by using a special generating 
equipment, a sufficient number of voltages can be obtained 



Digitized by VjOOQ IC 



§16 



DIRECT-CURRENT MOTORS 



61 



to give the necessary range in speed. Fig. 45 shows a 
multivoltage system that is well suited for the operation of 
machine tools. A, B, and C are three armatures genera- 
ting voltages of 110, 80, and 60 volts, respectively. The 
armatures may be mounted on the same shaft, and each pro- 
vided with its own field magnet; they may be three separate 
machines, or what is more usual, A may be a standard 
110-volt dynamo and B and C the commutators of a double- 
voltage dynamo, i. e., a dynamo having a single field and 
an armature provided with two windings connected to 
two separate commutators. The Bullock Electric Manu- 
facturing Company make apparatus for this system, and 
recommend the above voltages as suitable for machine-shop 



I/O Wits, 



BOIMhs. 



eowffs 




MM 



PIG. 45 

operation. The dynamos excite their own fields, and the 
armatures are connected in series, as shown. Four wires 7, 
2, Sy If are run to each motor, and the pressures obtained 
between the various lines are as indicated. Ordinary shunt- 
wound motors may be used, their fields being excited by 
connecting them across any pair of the wires, say, across 1 
and ^ or i and ^ since 110 and 250 are standard voltages. 
By' means of a suitable controller, tlie armature terminals 
can be connected between any pair of wires, so that the 
applied voltage can be any one of the following: 60, 80, 110, 
140, 190, 250. This will give six different speeds, and the 
speed corresponding to each voltage will remain constant 
no matter how the load on the motor may vary. If 



Digitized by VjOOQ IC 



63 DIRECT-CURRENT MOTORS §15 

intermediate speeds are desired, they can be obtained by 
means of a rheostat in the motor field. 

75, Fig. 46 shows a multi voltage system with a special 
generating outfit consisting of the two armatures A and B 
mounted on a common shaft. Armature A has two wind- 
ings, one of which generates 60 volts and the other 80 volts. 
Z> generates 110 volts. The armature of the motor is con- 
nected to a controller, represented diagrammatically in the 



speed Regufaffftg- 
fiheosfat 
PlO. 46 

figure by the two switches 5, 5', by means of which the 
armature can be connected across any pair of lines. The 
field is excited from the 110-volt mains and has a rheostat 
in series with it to admit of intermediate changes in speed. 

76, Fig. 47 shows a multivoltage system where a main 
110-volt generator A supplies current for lighting purposes 
and constant-speed motors. In order to take care of the 
variable-speed load, a motor generator set B C\^ installed. 



Digitized by VjOOQ IC 



§15 



DIRECT-CURRENT MOTORS 



63 



It is usually sufficient for this set to, have a capacity 
about 20 to 25 per cent, of the rated horsepower of the 
variable-speed motors, because most of the motors are, as 
a rule, operated at slow speed for a portion of the time 
only. Motor B drives the double-voltage generator C and 
thus renders available six different voltages. An ordinary 




pio.4r 

110-volt distributing system can therefore be converted 
into a multivoltage system by the addition of a motor 
generator set as described. Fig. 48 shows another arrange- 
ment, where the main generator is a 220-volt machine and 
the voltage is subdivided by a balancing set A B connected 
as shown. 

77, Fig. 49 shows a method of operating motors from 
an Edison three-wire system that is sometimes used where 
high and low speed are desired, but where a large number 
of different speeds are not necessary. It is, in fact, a- 
simple multivoltage system, and is very convenient for 

44—21 



Digitized by VjOOQ IC 



64 



DIRECT-CURRENT MOTORS 



§15 



the operation of printing presses, machine tools, etc. The 
armature is connected across 110 volts for low speed, 
and 220 volts for high speed. A certain range of inter- 
mediate speeds can be obtained by using a rheostat F in 
series with the field. A is the main switch, which is left 
closed during the time the motor is in operation. B is 
a double-throw switch; when thrown up it connects the 
armature across 110 volts, and when thrown down, across. 




PIO. 48 



Fio. 49 



220 volts. The field is excited from 220 volts as soon as 
switch A is closed, and remains excited no matter what the 
position of switch B may be; a starting rheostat E is 
placed in the armature circuit. It will be seen that when 
switch A is opened, the field circuit is broken, and in order 
to avoid an inductive discharge, lamps / may be connected 
across the field. In some cases a small auxiliary switch 
operated by the starting-box lever is connected in $erie9 



Digitized by VjOOQ IC 



§ 15 DIRECT-CURRENT MOTORS 65 

with these lamps, so that they do not burn until the starting- 
box lever moves from the on-position when* the circuit is 
opened at A. Sometimes an automatic release is provided 
on the field-regulating rheostat /% so as to make sure that 
the motor will always be started with full field strength. 

With two line voltages, it is possible to obtain a speed 
variation of from 1 to 6 by using resistance in the field. 
Where this system is regularly applied, the motor is con- 
nected to a controller that inserts the field resistance and 
changes over the armature from one voltage to the other, 
so that the speed may be changed smoothly and gradually. 
Of course, a regular controller constructed on similar lines 
to a street-railway controller is preferable to the arrange- 
ment of switches and rheostats shown in Fig. 49, and is 
more easily operated, but Fig. 49 shows a useful plan of 
connections in case a regular controller is not at hand. 



TBASBB SYSTEM OF CONTROI- 

78. The so-called booster-teaser system of control has 
been developed by the Bullock Electric Manufacturing 
Company, its special object being the control of large print- 
ing presses. These presses have to be run very slow during 
the process of making ready ^ and it is necessary to have 
some system of control that will allow the press to be 
moved very -slowly, perhaps only a few inches at a time, 
and also allow it to be run at any intermediate speed or 
at full speed as desired. 

79, Fig. 50 will serve to illustrate the general features 
of this method. A is the motor provided with a shunt 
field /and a series-field g, -5 is a small dynamo, or booster^ 
driven by a shunt motor C, B and C are mounted on the 
same base and are connected by a shaft. In series with A 
is a small amount of resistance d^ and an adjustable 
resistance e is placed in series with C, The booster gener- 
ator, in addition to its shunt field winding, not shown in 
the figure, is provided with a powerful series-winding k 



Digitized by VjOOQ IC 



66 



DIRECT-CURRENT MOTORS 



§15 



that causes the generator voltage to increase as the current 
delivered by it increases. 

Assume that the main circuit through the motor is opemd 
at / and that it is desired to operate motor yl at a very low 
speed and yet make it exert a powerful torque. Motor C is 
first started, using r as a starting resistance. The current 
for this motor, after passing through C, passes through A 
and thence to the line. The voltage E generated by B is 
applied to A^ and consequently the speed at which A will 
run and the amount of current, and hence the torque, can 
be adjusted by varying the voltage of B, Motor A can 
therefore be supplied with a large current at low voltage, 

Line + 




PIO. 50 

thus giving a large torque at low speed, without drawing a 
large current from the mains, because the motor generator 
set C B transforms the high-voltage current supplied to C 
to a low- voltage current delivered by B. For high speeds, 
the teaser can be dropped and the main motor A supplied 
with current through the resistance d with very little loss 
in efficiency, since at the high speeds d would be either cut 
out altogether or the amount included in this circuit would 
be small. The booster generator has an adjustable shunt h 
across its series-coils, so that their effect can be varied. 
The regulation of the booster voltage and the various con- 
nections necessary to operate the booster in connection with 
the main motor, together with the cutting in and out of 
resistance, are effected by a drum type of controller similar 
in general construction to a railway controller. 



Digitized by VjOOQ IC 



$15 



DIRECT-CURRENT MOTORS 



61 



CONTROL. BY VARIATION OF FTELB RELUCTANCE 

80. As previously stated, speed control by variation of 
the field exciting current is limited in range, the exact 
range depending on the design of the motor, and with ordi- 
nary motors, not exceeding 25 to 33 per cent. Motors are 
built in which the field strength is varied by leaving the 
exciting current constant and varying the reluctance of 
the. magnetic circuit to obtain different degrees of field 
strength. This method, as originated by Mr. F. A. John-^ 
son, is used in the Stow motors, and by means of it a speed 
variation of 126 per cent, can be obtained. At first glance' 




PlO. 61 



it would appear that a variation in field strength by vari- 
ation of the reluctance would have no advantage over a 
variation by a change in the exciting current. The objec- 
tion to the latter is that a considerable weakening of the 
field by cutting down the exciting current interferes with 
the commutation, whereas by the former method the field 
can be greatly weakened and a suitable commutating fringe 
maintained so that the motor will operate without sparking. 
Fig. 51 shows the peculiar style of pole piece used on the 
Stow multipolar motors; ♦ {a) shows the external appearance 



♦ G. Frederick Packard, Transactions American Institute Electrical 
Engineers, November, 1902. 



Digitized by VjOOQ IC 



68 DIRECT-CURRENT MOTORS §15 

and (i) a vertical section. The pole piece consists of 
two parts — an outer shell a and a movable core or plunger 6 
that can be moved up or down by means of the screw c; 
d is the yoke of the magnet frame, and it is evident that 
when b is at its lowest position, the pole piece offers the 
minimum reluctance to the magnetic flux, because the pole 
piece is then practically solid. The speed will therefore be 
a minimum in this position of the core. As the plunger is 
drawn up, the flux is compelled to reach the armature by 
way of the thin shell ^, which has a high reluctance and 
therefore cuts down the flux. Openings are cut in the shell, 
as shown at g^ and when the core is drawn up, practically 
the whole flux has to pass into the armature by way of the 
pole tips r,/, thus providing a good commutating fringe and 
preventing sparking. As the core is drawn up, it is evident 
also that there is considerable reluctance placed in the mag- 
netic path through which the cross-magnetizing turns act, 
and this also tends to secure sparkless operation. In the 
multipolar motors all the cores are connected by suitable 
gearing, so that they can be moved together. This special 
method of field control gives a much wider variation than the 
ordinary method, but it makes the construction of the motor 
somewhat expensive. 



DESIGN OF DrBECT-CURRENT MOTORS 

81. Since direct-current motors are constructed in the 
same way as direct-current dynamos, and the same require- 
ments for high efficiency apply to both, it follows that the 
machine that makes a good dynamo will, in general, operate 
well as a motor. On the other hand, some machines that 
operate fairly well as motors will not operate well as 
dynamos. For example, some small shunt motors and series- 
motors will not operate when driven as dynamos, because 
they are not capable of exciting their fields; whereas, when 
they are run as motors their fields are excited by current 
from the mains, and they are therefore capable of operation. 
Also, it is possible that a machine unstable in its operation 



Digitized by VjOOQ IC 



§15 DIRECT-CURRENT MOTORS 69 

as a shunt dynamo might operate all right as a motor, 
because in the latter case its fields are excited from the 
mains. In designing continuous-current motors, therefore, 
we may determine what the output of the motor would be if 
run at the required speed as a dynamo, and then design the 
motor as if it were a dynamo, making use of the various 
rules already given in connection with continuous-current 
dynamo design. 



DBTERMTN'ATION OF OUTPUT 

82. Suppose a given machine be run as a dynamo; the 
total electrical power developed in the armature will be equal 
to the power delivered at the terminals plus the loss due to 
armature resistance, and the loss in the field or the power 
delivered will be the total power developed in the armature 
multiplied by the electrical efficiency. When the machine 
is operated as a motor, the electrical energy developed in 
the armature will be the total electrical energy supplied 
from the mains less the loss due to field and armature 
resistances, or it will be the total energy supplied multiplied 
by the electrical efficiency of the motor. If E^ is the 
counter E. M. F. of the motor, / the current flowing through 
the armature, and E the E. M. F. between the mains, we 
have for a series motor 

total energy supplied from mains = £ / (15) 

because in the case of a series motor the current in both 
armature and field is /. Also, 

energy developed in armature = I E^ (16) 

and 

IE E 
electrical efficiency = , ^ = -^ (17) 

For a shunt motor, we have 

total energy supplied = IE '\- iE (18) 



Digitized by VjOOQIC 



70 DIRECT-CURRENT MOTORS §16 

where / is the current in the shunt field, and 

energy developed = lE^ 

IE 
electrical efficiency = ^ .^ *" . (19) 

In the case of a dynamo, the total electrical energy devel- 
oped in the armature is less than the total energy supplied 
at the pulley by the amount of the losses due to friction, 
hysteresis, and eddy currents. In a motor the useful output 
at the pulley is equal to the total electrical energy in the 
armature less the above losses, or is the total electrical 
energy generated multiplied by the efficiency of conversion 
of the motor. 

83, From the foregoing, it will be seen that if we know 
approximately the values of these efficiencies for the size of 
motor that we wish to design, we can calculate what the out- 
put of the motor would be if run as a dynamo, and then 



Pio. m 



proceed to design it as if it were a dynamo. Fig. 52 shows 
the approximate values of these efficiencies for motors up to 
160 horsepower. The upper curve A gives the electrical 



Digitized by VjOOQIC 



§15 DIRECT-CURRENT MOTORS 71 

efficiency at full load, that is, the ratio of electrical energy 
developed in motor armature to input, and the lower curve B 
gives the efficiency of conversion or the ratio of power devel- 
oped at pulley to electrical energy developed in armature. 



DESIOK OF 10-HOR8EPOWER SHITNT MOTOB 

84. We will suppose, for example, that it is desired to 
design a 10-horsepower shunt motor to operate on a 220- volt 
circuit and run at a speed of 1,000 revolutions per minute 
at full load. The field takes 3 per cent, of the total electri- 
cal input. It is required to find the current capacity of the 
armature of the corresponding dynamo, and also the voltage 
that it must generate when run at the above speed, so that 
we may proceed to design the motor as if it were a dynamo. 
The efficiency of conversion of a machine of this size is, 
according to Fig. 52, about .88, and the electrical efficiency 
is about .89. In order, then, to get a useful output of 

10 X 746 
10 horsepower, the input must be —^ ~~ = 9,525 watts. 

The line pressure is 220 volts; hence, the total current taken 

at full load is ' = 43.3 amperes, nearly. Of this current 

input 3 per cent., or about 1.3 amperes, flows around the 

field, so that the armature current at full load would be 

about 42 amperes. The total electrical energy in the arma- 

10 X 746 
ture is -^ — = 8,477 watts; hence, the voltage generated 

8 477 
in the armature must be * = 201.8 volts. In order, 

therefore, to obtain a motor that will deliver 10 horsepower, 
we must design a dynamo of which the armature has a 
current-carrying capacity of 42 amperes and that will gener- 
ate 202 volts, nearly, when run at a speed of 1,000 revo- 
lutions per minute. When this machine is run as a motor, 
the speed at no load will be slightly over 1,000 revolutions 
per minute, because the counter E. M. F. generated will 
then be nearly 220 volts. The shunt field would be designed 



Digitized by VjOOQ IC 



72 DIRECT-CURRENT MOTORS § 15 

for a current of 1.3 amperes, and the winding would be 
calculated in the same way as the shunt winding for a 
dynamo, except that no allowance would be made for a field 
rheostat, the winding being designed for connection direqtly 
to the 220-volt mains. 

85. The output of the dynamo corresponding to a series 
motor is determined in much the same way as that for a 
shunt motor, formulas 15, 16, and 17 being used. Of 
course, in a series motor the field winding must be capable 
of carrying the full-load current, but the efficiency, etc. for 
the two types of motor of the same output should be about 
the same. 

DE8IGX OF lO-HORSBPOWER SERIES MOTOR 

86. Suppose it were desired to design a lO-horsepower 
series motor to operate on 220-volt constant-potential mains. 
We will take the loss in the field as 3 per cent., as before, and 
calculate the current and voltage output of the correspond- 
ing dynamo accordingly. The efficiencies will be the same 
as before, that is, electrical efficiency = .89 and efficiency of 
conversion = .88. The total input will be 9,525 watts, or 
43.3 amperes. In this case, however, the armature must be 
designed for 43.3 amperes, instead of 42, as before. The 
total electrical energy in the armature will be 8,477 watts, 

8 477 
as in the last case, and the counter E. M. F. will be ' 

4<5.o 

= 195.8. The dynamo must therefore have an armature 

wound to deliver 43.3 amperes at 195.8 volts. The voltage 

generated by the armature is less in this case than with 

the shunt motor, because a portion of the line E. M. F. E is 

used up in forcing the current through the field coil. The 

speed of the motor would vary with the load, since the field 

magnetization varies with the load. 

87. It is especially important, in connection with the 
electrical side of motor design, to see that the field is very 
strong compared with the armature, in order to minimize the 
shifting of the brushes with the load. This is especially 



Digitized by VjOOQ IC 



§15 



DIRECT-CURRENT MOTORS 



73 



necessary in the case of motors, because they are liable to 
run under large and sudden fluctuations in load, and in many 
cases are frequently reversed, thus rendering a fixed point 
of commutation extremely necessary. 



MECHAXICAXi DESIGN 



STATIOXARY MOTOBS 

88. The general mechanical design of stationary motors 
is much the same as that of stationary dynamos ; in fact, in 
many cases the same castings, armature disks, etc. are 
used for both. Both series motors and shunt motors are built 
in the same way, the only difference being in the field winding. 
The parts of a motor should be made as simple as possible, 
since motors do not generally get the same amount of care 
as dynamos. Carbon brushes are used almost exclusively, as 
they tend to keep down sparking with changes in the load. 

89. For many classes of work it is desirable to have a 
motor that is partially enclosed. If the motor is in an excep- 
tionally dusty or dirty 
location, it should be 
wholly enclosed. It 
must be remembered, 
however, that the 
more a motor is en- 
closed the less are the 
facilities for the dissi- 
pation of heat, and 
for a given allowable 
temperature rise, a 
motor fully enclosed 
would have a con- 
siderably smaller 
capacity than the 
same motor open. ^®* * 

Fig. 53 shows a 10-horsepower motor of the semienclosed 




Digitized by VjOOQIC 



74 DIRECT-CURRENT MOTORS %15 

type made by the General Electric Company; it represents 
a style of construction very largely used for motors of 
moderate output. The bearing shields A^B are bolted to 
the field frame by means of four bolts, and by changing the 
position of the end shields, the motor can be attached to a 
side wall or suspended from a ceiling. If desired, the 
openings in the end frames can be closed by cast-iron 
covers, thus changing the motor into one that is wholly 
enclosed. It is better not to enclose a motor altogether 
unless it is absolutely necessary on account of dirt or flying 
particles. Fig. 64 shows a motor of similar type provided 
with back gearing in order to secure a slow speed. The end 
shields are provided with bearings that carry the secondary 



PlO. 64 

shaft, and the gears are protected by a gear-case that is 
partly filled with oil. In the figure, the upper half of the 
gear-case is removed. Motors of large size are usually con- 
structed on the same lines as generators; in fact, the same 
castings are frequently used for both. It will not be neces- 
sary, therefore, to give further illustrations. Street-rail- 
way motors form a class almost by themselves and detailed 
descriptions of them will be found in Electric Railways. 



Digitized by VjOOQIC 



§15 DIRECT-CURRENT MOTORS 76 



CARE A^TD OPERATTION OF DYNAMOS 
AKD MOTORS 

90, The following gives briefly some of the more 
important points relating to the care and operation of 
direct-current machines. As a rule, alternating-current 
machinery, described later on, does not require as close 
attention as direct-current, because on many alternating- 
current machines there is no commutator to give trouble. 
Space does not permit the consideration of all the troubles 
or faults that may arise in connection with direct-current 
machines, so that only the more common and important 
ones will here be taken up. 



GEKERAIi €ABE OF MACHINES 

91. The dynamo or motor, and all devices connected 
with its operation or regulation, should be kept perfectly 
clean. No copper or carbon dust should be allowed to 
accumulate to cause breakdowns in insulation. The oil 
gauges and grooves should be kept in working order and 
the oil in the wells should be renewed at regular intervals. 
The brushes should be- kept clean. They should be set and 
trimmed to fit the commutator, and if copper brushes are 
used, they should be taken out once in a while, whenever 
they become clogged, and dipped in gasoline to cleanse 
them. New carbon brushes should be sandpapered to fit 
the commutator, by sliding a piece of sandpaper back and 
forth between them and the commutator. Do not use 
emery paper for cleaning the commutator, as emery is more 
or less of a conductor and may cause short-circuiting 
between the bars; also small pieces of emery become lodged 
in the brushes and scratch the commutator. 



Digitized by VjOOQIC 



76 DIRECT-CURRENT MOTORS §15 

Oil should be used very sparingly, if at all, on the com- 
mutator; to lubricate it, put a film of vaseline on a canvas 
cloth, fold the cloth once, and let the commutator get only 
what goes through the pores. Never allow a loose article 
of any kind to lie on any part of a machine. 




BBUSHBS 

92. On direct-current machines, the brushes and com- 
mutator require, perhaps, more attention than all the other 

parts of the machine put 
together. Brushes are of 
two kinds: radial and tan- 
gentiaL Radial bmslies 
point straight at the center of 
the commutator; their direc- 
tion is parallel to the radius, 
as shown in Fig. 55 (a), Tan- 
grential brushes. Fig. 55 (^), 
are frequently made of cop- 
per and are found, as a rule, 
on lighting machines. Radial 
^®* * brushes are made of carbon. 

Carbon radial brushes are always used on machines that 
must admit of being reversed in rotation. With carbon 
brushes, the commutator takes on a dark-chocolate polish 
and the brushes emit a squeaking noise at starting or 
stopping. 

93, Carbon brushes are made in several grades of hard- 
ness, adapted to different conditions of working and differ- 
ent kinds of commutators. In stationary direct-current 
work, soft carbon is used ; on street-car work, hard carbon. 

The proper pressure for a brush depends on the material 
and condition of the commutator and onthe material of the 
brush itself. A copper brush does not, as a rule, call for as 
much tension as a carbon brush, and soft carbon will run 



Digitized by VjOOQ IC 



§15 DIRECT-CURRENT MOTORS 77 

with less tension than hard carbon; a rough commutator 
needs more brush tension than a smooth one; for given 
brush contact, large currents call for more pressure than 
small ones. Finally, where there are several brushes in 
each holder, the tension must be the same on all, so that 
they will all take about the same current. This tension 
should not be great enough to wear out the commutator 
unnecessarily or cause an unnecessary amount of friction. 

If the contact between the brush and the commutator is 
loose, the contact resistance will be high and heating will 
result. On the other hand, increasing the pressure beyond 
a certain amount results in very little reduction of resist- 
ance, but greatly increases the friction. For stationary 
work, a pressure of 2 to 2^ pounds per square inch of brush 
contact surface should be sufficient. For railway work, the 
pressure has to be much heavier on account of the jarring 
to which the motor is subjected. 

94. One weakness of carbon brushes is that they, at 
times, stick in the holder so that the tension spring is not 
strong enough to force down the brush to its place, and 
even if it does force it down, the pressure on the com- 
mutator will be too light. This very serious fault may be 
due to either of two causes: lack of uniformity in the thick- 
ness of the brush, or an excess of paraffin in the brush. If 
a brush is thicker at one end than it is at the other, it may 
go into the holder freely if put in thin end first, and might 
not go in at all on the thick end. The result of this is that 
as soon as the brush wears down to a point where the thick 
end enters the holder it sticks. On the other hand, the 
fault may be due to a nick or burr in the brush holder 
itself. 

The second source of trouble — an excess of paraffin in the 
brush — is accounted for as follows: If a carbon brush is 
snugly fitted into the holder so that it slides back and forth 
freely, but without any clearance, while the holder and 
brush are cold, as soon as they become warm, the paraffin, 
which is mixed with the carbon for lubricating purposes, 



Digitized by VjOOQ IC 



78 DIRECT-CURRENT MOTORS §16 

oozes out, forms a paste with whatever carbon or copper 
dust there may be present, and causes the brush to stick. 
It is very essential that the brushes should be kept clean 
and trimmed to fit the commutator. 



THB COMMUTATOB 

95. The commutator is the most sensitive part of a 
machine, and its faults are liable to develop more quickly 
than those of any other part. When a commutator is in 
the best possible condition, it becomes a dark-chocolate 
color, is smooth, or glazed, to the touch, and causes the 
brushes, if of carbon, to emit a characteristic, squeaky 
noise when the machine is turning slowly. Under no cir- 
cumstances should any weight be allowed to rest on the 
commutator, nor should it be caught with a sling when the 
armature is being lifted. To secure the best results, 
the brush holders should be set as close to the commutator 
as possible, so as to do away with chattering. 

96, If a dynamo or motor is not overloaded too much, 
if the brushes are set properly, and if the commutator is 
made of the proper material, it should seldom get rough. 
As a rule, sandpaper and emery cloth are used around 
machines much more than they should be. For ordinary 
roughness of the commutator, due to some temporary 
abnormal condition, it is well enough to use sandpaper, but 
for chronic roughness some more permanent treatment 
must be applied. 

If a commutator, when it is built, is not properly baked 
or screwed down after it is baked, it is liable to bulge out in 
the course of time under the action of the heat due to its 
normal load and the centrifugal force, or it may develop 
loose bars. In the case of the bulging of one side, sand- 
paper will not do any good. The best thing to do with a 
commutator that bulges badly is to take it off, bake it so as 
to loosen the insulation, tighten it up well, and turn it off 



Digitized by VjOOQ IC 



§15 DIRECT-CURRENT MOTORS 79 

in the lathe. For ordinary curvatures of surface, that is, 
unerenness due to wear, it is customary to set up a tool post 
and slide rest on the bedplate of the machine itself and turn 
off the commutator in position. 

A narrow scratch or several of them all around the 
commutator means that there are particles of hard foreign 
matter under one or more of the brushes. A broad scratch 
around the bearing surface of the commutator probably 
means that one of the brush holders has been set too close 
or has become loose and slipped down to a point where it 
touches the bars. 

97. 'High. Bars. — A metallic click emitted twice, four 
times, or six times (according to the number of brush 
holders in use) per revolution indicates a high bar in the 
commutator ; in such a case, the brushes will be seen to 
jump a little when the high bar passes under them. A 
high bar may be due either to a loose bar working out or to 
the fact that orte bar is much harder than any of its neigh- 
bors, and therefore does not wear down at the same rate. 
If the high bar seems to be firm under a blow from a 
hammer, it will be safe to take it down with a file while the 
armature stands still ; but if the hammer test proves the bar 
to be floating, it is a serious matter, and nothing short of a 
regular repair job will give satisfaction. In testing a bar or 
bars with the hammer, care must be taken not to nick or 
dent the commutator, as such a defacement will cause the 
high-bar click to be emitted and it will be misleading. 

98, liovr Bars. — A fault very much akin to a high bar 
is the low bar, which gives forth very much the same sound, 
but the brushes drop, instead of rising, as the fault passes 
under them. The low-bar trouble may be due to any of 
several causes: it may be due to the commutator having 
received a severe blow in the course of handling; to one or 
more bars being of poorer material than the rest ; or it may 
be due to the gradual eating away of the bar on account of 
sparking at that particular place. A high bar can be 

44—22 



Digitized by VjOOQ IC 



80 DIRECT-CURRENT MOTORS §15 

removed by filing down or turning down that bar alone to 
the level of the others, but to get rid of a low bar, the rest 
of the commutator must be turned down to its level. 

99, Of course, the most serious condition is to have a 
commutator that is poorly made and of poor material. If the 
mica and copper used are not of the proper relative hard- 
ness, one will wear down faster than the other, leaving the 
surface of the commutator a succession of ribs. If the 
mica is too soft, it will pit out between the bars, leaving a 
trough to fill up with carbon dust and in a degree short- 
circuit the neighboring armature coils. If the mica bodies 
are too hard or too thick, the bars will wear in ruts and call 
for frequent turning down. The brushes should be set so 
that, with a slight end play of the armature, the whole 
commutator surface will be utilized. 



THE ARMATURE 

100. Armatures should be handled with great care, as 
it is an easy matter to bruise a coil or a lead, and this may 
not be noticed until the machine is started up. All arma- 
tures should be supported by their shafts as much as pos- 
sible. When lying on the floor, they should lie on padding 
of some sort. Extra armatures not in use should be kept 
housed in a dry place. 

101, Heating: of Armatures. — An armature should 
run without excessive heating; if it heats so as to give off 
an odor, any of several things may be the matter. It may 
be damp — a condition that, as a rule, is shown by steaming, 
but which can be better determined by measuring the insu- 
lation to the shaft with a voltmeter. If low insulation is 
indicated, the armature should be baked, either in an oven 
or by means of a current passed through it in series with 
a lamp bank or water resistance. The baking current 
should not exceed the full-load current of the machine. 



Digitized by VjOOQ IC 



§15 DIRECT-CURRENT MOTORS 81 

In applying the current to the armature, be sure that the 
series-field, if the machine has one, is not included in the 
circuit, and that the shunt field is broken; for if either 
field is on, the machine may start up as a motor. 

10!3« If, instead of the whole armature running hot, the 
heat is confined to one or two coils, the indications are that 
there is a short circuit either in a coil or between the two 
commutator bars to which the ends of the coil connect. 
Such a short-circuited coil run in a fully excited field will 
soon burn out. A short circuit of this kind can be readily 
detected by holding an iron nail or a pocket knife up to the 
head of the armature while it is running in a field; any 
existing short circuits in the coils or commutator will cause 
the piece of metal to vibrate very perceptibly. One or more 
coil connections reversed on one side of the armature will, 
on a dynamo, cause a local current to flow from the strong 
half to the weaker half, and thereby cause all the coils to 
heat more than they should. On a motor, the effect is 
to decrease its counter E. M. F. so that it will take more 
current under given conditions, while the side containing 
the reversed coils will be hotter than the other side. A test 
for locating reversed coils will be descril^ed later. 

103. One very peculiar fault to which armatures are 
liable is known as a flylngf cross. This is due to a loose 
wire that only gives trouble when the armature is in motion. 
The loose wire may either be broken, it may have a loose 
connection, or it may have defective insulation that allows 
it to come in contact with other wires as soon as the arma- 
ture comes up to speed. In any case, the fault gives no 
trouble as long as the armature is at rest, but as soon as it 
gets up to speed, the centrifugal force throws the loose wire 
out of place and causes the brushes to flash. 

104« Overloaded Armatures. — One of the most com- 
mon causes of general trouble and heating in an armature is 
OYerload ; this may be due to ignorance or neglect or to an 



Digitized by VjOOQIC 



8a DIRECT-CURRENT MOTORS §15 

error in the instrument that measures the load. There is a 
great tendency on the part of owners to gradually increase 
the load on a machine until it may be doing about twice the 
work it is intended to do. If the machine is a dynamo, lamps 
are added to its load one or two at a time; in this way it is 
an easy matter to overload a dynamo without intending to. 
If the machine is a motor, small devices may be put on it, 
one at a time, until an overload is the result. Where 
machines are running together in multiple, one may be 
taking more than its share of the load, due to poor equali- 
zation. Ammeters sometimes get out of order, read incor- 
rectly, or stick. 



fii:lj>-cx>ili bbfects 

105* Fields, like armatures, are subject to defects of 
various kinds; they are liable to open circuits, short circuits, 
and grounds. In the case of a series dynamo, an open cir- 
cuit in the field will render it totally incapable of operating 
either as a motor or a dynamo, but no harm can come to the 
machine itself, unless the fault should take place while the 
current was on, in which case it would be apt to burn a hole 
in whatever happened to be around the break. 

106* On a shunt machine, the amount of trouble caused 
by the opening of a field coil depends on whether the 
machine is a motor or a dynamo; and if a dynamo, on 
whether it runs alone or in multiple with other dynamos. 
If the machine is an isolated dynamo, a break in the field 
coil can do no further damage than to prevent the machine 
from generating. If, however, the dynamo is in multiple 
with other dynamos on the same load, the result of such a 
fault will be that the other machines will send a large rush 
of current back through the faulty machine and cause, 
practically, a short circuit. Breaking the field of a shunt 
motor destroys its counter E. M. F., so that there is no 
opposition to the line E. M. F. and there is practically a 
short circuit through the armature. 



Digitized by VjOOQ IC 



815 t)IRECT-CURRENT MOTORS 83 

107. On a compound-wound motor connected differ- 
entially, as soon as the shunt field breaks, the series-coils 
opposed to it bring the motor to a stop and start it up in 
the opposite direction. A compound-wound motor will, il 
the starting current is large enough, start up on the field 
provided by the series-turns. 

108. The effect of the reversal of a single field coil on a 
machine depends on how many field coils the machine has; 
in other words, the more poles a machine has, the less will 
be the effect of an irregularity in one of them. If the 
machine has only two coils and one of them is incorrectly 
connected, the machine will not start as a motor and it will 
not generate as a dynamo. If there are four field coils, it 
will take a heavy current to start it as a motor, and the 
brushes will spark even while the motor is starting. 

109. Short Circuits In Field. — The action of a short 
circuit in a field coil depends on the kind of machine and the 
manner in which the field coils are connected. Consider a 
shunt dynamo with four field coils; if the coils are in series, 
so that the voltage across each coil is only one-fourth of the 
total voltage, a short circuit in a single coil will cause it to 
run comparatively cool, while the remaining coils will get 
unusually warm. The cutting out of a single coil in four 
will reduce the resistance of the field circuit so that more 
current will flow through the remaining coils. 

110. Moisture In Field Ck>lls. — Moisture in field coils 
will cause them to heat. Before putting such fields into 
actual service, they should be baked out, either with the 
current or in an oven. The best way to locate a short- 
circuited field is to measure the resistance of the field 
suspected and compare it with that of a standard field 
of the same kind. A short-circuited shunt field can be 
located by short-circuiting or cutting out one field coil 
at a time and measuring the open-circuit voltage of the 
dynamo. 



Digitized by VjOOQ IC 



84 DIRECT-CURRENT MOTORS §15 



REASON^S FOR A BYNAMO PAIL.TNG TO GEN1BRATB 

111. lioss of Residual Magrnetism. — Among the many 
causes that may make a dynamo fail to generate, loss of 
residual magnetism is often one of the most troublesome. 
As a rule, dynamos leaving the factory retain enough 
residual magnetism to start on, but there are several ways 
in which they can los^e it. Some dynamos never lose their 
residual magnetism, or diargre, as it is called, while others 
seem to have a weakness for doing so. 

113, Where a dynamo has lost its charge, the pole 
pieces will have little or no attraction for a piece of soft 
iron. Series dynamos seldom lose their charge so entirely 
that they will fail to pick up a field on short circuit. 
Where a compound-wound dynamo refuses to pick up a 
field with its shunt field, it can often be made to pick up by 
disconnecting the shunt coils and short-circuiting the 
machine through a small fuse. Machines can in some cases 
be made to pick up a field by simply short-circuiting the 
armature by holding a piece of copper wire across the 
brushes or by rocking the brushes back from their neutral 
position. 

If none of these expedients produce the desired result, 
the fields must be recharged from an outside source. If the 
dynamo runs in multiple with other dynamos, this is an 
easy matter; it is only necessary to lift the brushes or dis- 
connect one of the brush-holder cables on the dead machine 
and throw in the main-line switch, the same as if the 
machine were going into service with the others. The 
fields will then take a charge from the line and their 
polarity will be correct. If the dynamo does not run in 
multiple with another and there is a dynamo within wiring 
distance, disconnect the shunt field of the dead dynamo and 
connect it to the live circuit. If there are absolutely no 
other means available for charging, several cells of ordinary 
battery may be used. As a last resource, when all other 
available sources fail, connect the fields so as to obtain the 



Digitized by VjOOQ IC 



§15 DIRECT-CURRENT MOTORS 85 

least possible resistance, put them in series with the arma- 
ture through a small fuse, and speed the armature con- 
siderably above the normal rate. Very often a dynamo, 
instead of losing its residual magnetism, will acquire one of 
a reversed polarity, due, perhaps, to the same causes exer- 
cised to a greater degree. 

11 3. Wrong Connection of Field or Armature. 

Every dynamo requires that a certain relation exist 
between the connection of its field and armature and its 
direction of rotation, or it will refuse to generate. Suppose 
a dynamo to be generating; if its field or armature connec- 
tion (either, but not both) be reversed, it will be unable to 
generate; or if all the connections be left intact and the 
direction of rotation reversed, the machine will be rendered 
inert. Not only is it unable to generate with the wrong 
connections or rotation, but a short run under this condi- 
tion will render the machine unable to generate after the 
conditions are corrected, unless the field is recharged, 
because the effect is to destroy its residual magnetism. If, 
then, a dynamo fails to generate, and all other conditions 
are apparently correct, reverse the field terminals, and see 
if the machine will pick up. 

A shunt, or compound-wound, dynamo will not pick up it 
the shunt field circuit is open; the open circuit may be in the 
field itself, in the field rheostat, or in some of the wires or 
connections in the circuit. A careful inspection will gener- 
ally disclose any fault that may exist in a wire or connection. 
To find out if the rheostat is at fault, short-circuit it with 
a piece of copper wire; if the dynamo generates with the 
rheostat cut out, the fault is in the rheostat. To find out 
if the open circuit is in a field coil, use a test-lamp circuit 
or a magneto-bell to test the coils one at a time. Before 
doing so, be certain that all communication is cut off 
between the machine under test and the line, if there are 
other machines on the same circuit. A field circuit is some- 
times held open by a defective field switch that, to all 
appearances, is all right ; repeated burning will oxidize the 



Digitized by VjOOQ IC 



86 DIRECT-CURRENT MOTORS §16 

tip of the blade and make a non-conducting blister on it; 
the blister will not carry current and it will press the jaws 
of the switch apart so that only the blister touches, and so 
opens the circuit. Another trivial but common cause of 
open circuits is the blowing of fuses. 

An open circuit in the armature will interfere with the 
proper generation of the current, but such a fault, as a rule, 
announces its own occurrence in a very emphatic manner 
and does not, therefore, require to be looked for. 

Before attributing the failure of a dynamo to generate to 
any of the foregoing open-circuit causes, see that the brushes 
are on the commutator, the field switch closed, and the 
greater part of the field rheostat cut out. 

Always bear in mind that the E. M. F. generated when a 
machine is started up is very small, because the residual 
magnetism is weak. It may not require a complete open 
circuit in the field to prevent the machine picking up. A 
bad contact that might not interfere with the working of 
the machine when it is up to full voltage might be sufficient 
to prevent its picking up when started. A loose shunt wire 
in a binding post, or a dirty commutator, will introduce 
sufficient resistance to prevent the machine from operating. 
Trouble is very often experienced in making machines with 
carbon brushes pick up, especially if the brushes or commu- 
tator are at all greasy. If such is the case, thoroughly 
clean off the commutator, wipe off the ends of the brushes 
with benzine, and see that they make a good contact with 
the commutator surface. 

114. Short Circuits, — A short circuit on the line will 
make a shunt dynamo drop its field. With a short circuit 
on the line, a shunt dynamo will not, therefore, pick up its 
field. With a series-wound or compound-wound dynamo, a 
short circuit on the line increases its ability to generate, be- 
cause the fault is in series with the series-coils and a large 
current passes through them. A series dynamo, like a shunt 
dynamo, will not pick up if the field is short-circuited. A 
compound-wound dynamo will not pick up on open circuit 



Digitized by VjOOQ IC 



§16 DIRECT-CURRENT MOTORS 87 

if the shunt field is short-circuited; it will pick up with an 
open circuit in the main circuit, but will not hold its voltage 
under load if the series-coils are short-circuited. In some 
cases a shunt dynamo will not pick up on full load, as this 
realizes too nearly the condition of a short circuit; so that 
to be on the safe side, it is best to let the machine build up 
its field before closing the line switch. 

Short circuits within the dynamo itself generally give rise 
to indications that point out the location and nature of the 
fault. In any event, the first thing to find out is whether 
the fault is in the dynamo or out on the line; if it 
picks up its field when the line switch is opened, but fails to 
do so with it closed, the trouble is outside of the dynamo. 

115. Field Colls Opposed. — Failure to generate may 
be due to one or more field coils being incorrectly put on, 
or connected, so that they oppose each other. On a com- 
pound-wound dynamo, the reversal of a shunt-field coil will 
generally keep the dynamo from picking up on open circuit, 
unless the dynamo has more than four coils; the more coils 
it has, the less effect has the reversal of a single coil. The 
reversal of a series-coil is not felt until an attempt is made 
to load the machine ; the voltage will not come up to where 
it should for a given load, and the brushes are apt to spark 
on account of the weakening of the field. " 

116. liOw Speed. — No dynamo will pick up its field 
below a certain speed, but with the field once established, 
the machine will hold it at a much lower speed than that 
required to establish it. The speed at which a series- 
dynamo will pick up depends on the resistance of the 
external circuit. 

117» Among the causes of failure to generate not 
included in the foregoing are faults in the iron circuit, loose 
or open joints in the frame proper or between the pole 
pieces and the frame, and brushes not placed at the 
neutral position. Such imperfections also lower the maxi- 
mum voltage of the dynamo. 



Digitized by VjOOQ IC 



88 DIRECT-CURRENT MOTORS 816 



FATLiniE OF MOTOR TO START 

118. When a motor fails to start when the controlling 
switch is closed, any one of several things may be the 
matter. There may be an open circuit, a short circuit, 
a wrong connection, the power may be off the line, or 
the trouble may be purely mechanical. If the failure to 
start is due to an open circuit or to absence of power on 
the line, there will be no flash when the switch is closed 
and opened again. To tell if there is any power on the 
line, test with incandescent lamps or a voltmeter. If 
the fault is an open circuit, it may be found in any of 
the following places: Defective switch; broken wire or 
connection in the starting box; loose or open connection 
in some of the wiring; a piece of foreign matter under 
one brush; brush stuck in the holder or no brush in it 
at all; brush springs up; fuse blown; some wire, appar- 
ently all right, broken inside the insulation; or an open 
circuit in some part of the motor itself. If the trouble 
is due to a. short circuit, there will be a flash when the 
starting box is thrown off. 

119. Among the more common sources of short circuit 
are: short-circuited armature coils; short-circuited com- 
mutator; short-circuited field coils; field on a shunt or 
compound-wound motor connected so that tlie armature 
cuts out the field winding; brushes in the wrong position. 
If the armature coils or commutator are short-circuited, 
the machine may start and turn over part way and stop 
again. With a field coil short-circuited, the armature will 
start only under a heavy current, with accompanying spark- 
ing, and will acquire a high rate of speed. 

120. Wrong Connections of Shunt Field. — As pre- 
viously pointed out, it is an easy matter to confuse the 
connections of a shunt motor so that the shunt field will not 
be excited at starting or at least have a very small excita- 
tion, because of its being connected across the brushes. If 



Digitized by VjOOQ IC 



§16 DIRECT-CURRENT MOTORS 8d 

such a mistake is made, the motor will fail to start unless 
an excessive current is made to flow through the armature 
by cutting out the greater part of the starting rheostat, 
and even then the motor may not start unless the load is 
light. 



SPARKING 

121» Sparking at the brushes may be due to any of the 
following causes: Too much load; brushes improperly set; 
commutator rough or eccentric; high or low bars; brushes 
making poor contact; dirty brushes or commutator; too high 
speed; sprung armature shaft; low bearings; worn commu- 
tator; short -circuited or reversed armature coil; high-resist- 
ance brush; vibration; belt slipping; open-circuited armature; 
weak field; grounds, 

122. Too Much. lioad. — In this case the armature heats 
all over. The sparking may be lessened but nol stopped by 
shifting the brushes ahead on a dynamo and back on a 
motor. If the machine is a motor, the speed will be low ; 
if a dynamo, the voltage will be below the normal amount. 

123* Brushes Improperly Set. — Brushes may be out 
of their proper position in cither of two ways: they may be 
the right distance apart but too far one way or the other 
as a whole; this can, of course, be remedied by shifting the 
rocker-arm back and forth until the neutral point is found. 
The brushes may, as a whole, be central on the commutator, 
but too far apart or too close together. Such a fault must 
be remedied by adjusting the individual holders. 

124* Conn imitator Rough or Eccentric. — A commu- 
tator will become rough either as a result of abuse or as a 
result of bad selection of the copper and mica of which it is 
made. An eccentric commutator acts like a bent shaft and 
may be the result of faulty workmanship or the result of a 
hard blow. In either case it must be turned true, but 



Digitized by VjOOQ IC 



do DIRECT-CURRENT MOTORS § Ifi 

before doing it be certain that the commutator is at fault 
and not the shaft. 

125. Higli OP liO^w Bar. — A high or low commutator 
bar causes a poor contact between brush and commutator, 
and hence gives rise to sparking. 

126. Brushes Makings Poor Contact. — The brush may 
be stuck in the holder; the temper may be out of the tension 
spring; the brush hammer may rest on the side of the 
holder and not on the brush ; the brush may not fit the sur- 
face of the commutator; the holder may have shifted to the 
wrong angle. Tension springs should be paralleled by a 
conductor attached to the brush, so that the current will not 
flow through the springs. 

127. Dirty Brushes or Commutator. — Some carbon 
brushes are liable to give out paraffin when hot, which, 
getting on the commutator, insulates it in spots. A copper 
brush is apt to get clogged with oil, dust, and threads of 
waste (waste should never be used on a commutator). Dirty 
commutators, as a rule, are the result of using too soft a 
brush. 

128. Too Higrli Speed. — A machine is apt to spark if 
its speed is too high, because it interferes with the commu- 
tation. 

129. Sprung Armature Sliaft. — A sprung armature 
shaft causes the commutator to wabble, giving very much 
the same symptoms as an eccentric commutator. 

130. liOiiv- Bearings. — On some types of machine, 
excessive wear in the bearings throws the armature far 
enough out of center to distort the field and cause sparking. 

131. Worn Commutator. — When a commutator wears 
down below a certain point, even if otherwise in good con- 
dition, it seems inclined to spark in spite of everything that 
can be done. It may be because the brushes then span 



Digitized by VjOOQ IC 



§15 DIRECT-CURRENT MOTORS 91 

more bars, because the bars become thinner as they wear 
away, or it may be because an error in the angle of the 
holder increases with the distance from the commutator. 

132. Short-circuited or Beversed-Armature Coll. 

Either of these faults will cause a local current to flow, with 
the result that either a dynamo or a motor will require an 
unusual amount of power to run it even when unloaded. 
A motor will run with a jerky motion especially noticeable 
at low speeds, and the voltmeter connected to a dynamo 
will fluctuate. Such a fault may be due to a cross in the 
coil itself or contact between two commutator bars. In 
either case, unless the cross is removed, the coil will burn 
out. 

133. Hlgli-Reslstance Brusli. — Up to a certain point, 
high resistance in a carbon brush is a good feature, but it is 
possible to get the resistance so high that the brush will 
spark on account of its inability to carry the current at the 
contact surface. 

134. Vibration. — A shaky foundation will cause the 
whole machine to vibrate and will cause it to spark steadily, 
which fault can be remedied only by placing the machine on 
a firmer foundation. 

135. Belt Slipplngr. — A slipping belt will sometimes 
cause intermittent sparking because it subjects the machine 
to unusual variations in speed. 

136. Open-Circuited Armature. — By an open-cir- 
cuited armature is meant a break in one of the armature 
wires or its connections. Excessive current may burn off 
one of the wires or a bruise of some kind may nick a wire so 
that the normal load or less burns it off. A commutator 
may become loose and break off one or more leads. In any 
case there are two very characteristic symptoms of an open- 
circuited armature: a ball of fire runs around the commuta- 
tor and the mica is eaten from between the bars to which 



Digitized by VjOOQ IC 



92 DIRECT-CURRENT MOTORS §15 

the faulty coil is connected, the bars themselves become 
dark, pitted, and burned on the edges. Sometimes, on 
account of abuse, the armature throws solder and all the 
commutator connections become impaired. In such a case 
there are no actual open circuits, but there are poor con- 
tacts that result in making the commutator rough and 
black. 

137» Weak Field. — A weak field may be due to a loose 
joint in the iron circuit, to a short circuit in the field coils, 
to opposition of the field coils, or to the fact that heat has 
carbonized the insulation so that the current short-circuits 
through it. Any of these influences decrease the field 
strength, with the result that the starting power of the motor 
is decreased and the speed and current are increased. 
On a dynamo, the E. M. F. and the ability to pick up are 
decreased. 

138. Grounds. — On an ordinary circuit, a single ground 
has no effect, but two grounds can so take place that the 
whole or any part of the field or armature may be cut out; 
such a pair of grounds is nothing more nor less than a short 
circuit, and it falls under that head. 



TESTING FOR FAUIiTS 

139. Many of the defects that are liable to arise in con- 
nection with dynamos and motors are, of course, apparent 
from a mere inspection of the machine. Other defects, 
such as short-circuited or open-circuited field coils, short- 
circuited or open-circuited armature coils, etc., must be 
located by making tests. For tests of this kind, Weston or 
similar instruments are most convenient if they have the 
proper range for the work in hand. For measuring resist- 
ances, the iirop-of-[^otential method is generally most easily 
applied. This method consists in sending a known current 



Digitized by VjOOQ IC 



§ 15 DIRECT-CURRENT MOTORS 93 

through the resistance to be measured and noting the 
pressure between the terminals of the resistance; in other 
words, noting the pressure required to force the known cur- 
rent through the unknown resistance. If the resistance to 
be measured is very low, as, for example, an armature coil, 
the voltmeter must be capable of reading low and a milli- 
voltmeter (one reading to thousandths of a volt) will be 
best suited to the work. Of course, a good Wheatstone 
bridge may also be used for measuring resistances, but it is 
generally not as convenient to use around a station as the 
drop-of -potential method. 

140. Testing for Open-Circuited Field Colls, — If a 

machine does not pick up, it may be due to the absence of 
residual magnetism. If any residual magnetism is present, 
a voltmeter connected across the brushes will give a small 
deflection when the machine is run up to full speed, so that 
this point can easily be determined before a test is made 
for a broken field coil. Examination of the connections 
between the various coils will show if they are defective or 
loose; quite frequently the wire in the leads from the spools 
becomes broken at the point where the leads leave the 
spool, while the insulation remains intact, so that the break 
does not show. This may be detected by wiggling the 
leads. 

If the break is inside the winding of one of the coils, it 
can only be detected by testing each coil separately to see if 
its circuit is complete. This may be done with a Wheat- 
stone bridge or with a few cells of battery and a galvanom- 
eter. A low-reading Weston voltmeter makes a good 
galvanometer for this purpose. 

If the current from another dynamo can be obtained, the 
faulty spool may be detected by connecting the terminals of 
the field circuit to the terminals of the circuit of the other 
machine; no current will flow through if the circuit is 
broken, but if a voltmeter is connected across each single 
field coil in succession, it will show no deflection if the coil 
is continuous, because both poles of the voltmeter will be 



Digitized by VjOOQ IC 



94 



DIRECT-CURRENT MOTORS 



§X5 



connected to the same side of the dynamo circuit. If the 
coil has a break in it, one of its terminals will b.e connected 
to one side of the circuit and the other to the other side, so 
that a voltmeter connected between these terminals will 
show the full E. M. F. of that circuit. Consequently, 
when the voltmeter is connected across a spool and shows a 
considerable deflection, that spool has an open circuit which 
must be repaired before the dynamo can operate. 

This method of testing is represented by the diagram, 
Fig. 56; i, ^, 3, and .^ represent the field coils of a dynamo, 

there being a break in 
coil 2 at B. The ter- 
minals a and e are con- 
nected to the + and 
— terminals of a live cir- 
cuit. It will be seen that 
terminals a and b of coil 1 
are both connected to 
the + side of the circuit, 
and as there is no cur- 
rent flowing through the 
field circuit, there is no 
difference of potential 
between a and b\ there- 
fore, a voltmeter connected to a and b, as at V, will show 
no deflection. But terminal c of coil 2 is connected to the 
— side of the circuit ; so a voltmeter connected to b and r, 
as at F„ will show a deflection, and, in fact, will indicate 
the difference of potential between a and e. 

141. Short-circuited Field Coll.— It is evident that 
if the windings of a field coil become short-circuited either 
by wires coming in contact or by the insulation becoming 
carbonized, the defective coil will show a much lower resist- 
ance than it should. The drop of potential across the 
various field coils should be about the same for each coil, 
so that if one coil shows a much lower drop than the others, 
it indicates a short circuit of some kind. 




FIO. 56 



Digitized by VjOOQ IC 



§15 



DIRECT-CURRENT MOTORS 



96 



142. Test for Grounds Bet\veen Winding: and 
Frame. — After the machine has thoroughly warmed up, it 
should be tested for grounds or connections between the 
winding and the frame or armature core. This may best 
be done with a good high-resistance voltmeter, as follows: 
While the machine is running, connect one terminal of the 
voltmeter to one terminal of the dynamo, and the other 
terminal of the voltmeter to the frame of the machine, as 
represented in Fig. 57, where T and Z", are the terminals 




PlO. 57 

of the dynamo and V and F, two positions of the volt- 
meter, connected as described above. 

If in either position the voltmeter is deflected, it indicates 
that the field winding is grounded somewhere near the other 
terminal of the dynamo; that is, if the voltmeter at V 
shows a deflection, the machine is grounded near the ter- 
minal 7^,, and vice versa. If the needle shows a deflection 
in both positions, but seems to vibrate or tremble, the 
armature or commutator is probably grounded. If in either 
case the deflection does not amount to more than about -^ 
the total E. M. F. of the machine, the ground is not seri- 
ous, but if the deflection is much more than this, the wind- 
ings should be examined separately, the ground located, 
and, if possible, removed. 

44—23 



Digitized by VjOOQ IC 



96 



DIRECT-CURRENT MOTORS 



§15 




143. To locate the ground, each coil should be discon- 
nected from its neighbor (with the machine shut down, 
of course) and tested out by connecting one terminal of 

another dynamo (or of a live 
<i _ circuit) to the frame of the 
machine, care being taken 
to make a good contact 
with some bright surface, 
such as the end of the shaft 
or a bolt head; the other 
terminal of the coil to be 
tested is connected to the 
other line through a volt- 
meter, as represented in 
Fig. 58. 

^°- w Here C and C, represent 

the terminals of a live circuit, which should have a dif- 
erence of potential between them about equal to the 
E. M. F. of the machine when it is in operation, but not 
greater than the capacity of the voltmeter will allow of 
measuring. T and 7", represent the terminals of the 
dynamo, as before, and / and /, the terminals of the field 
coils, which have been disconnected from each other and 
from the dynamo terminals. One terminal C oi the circuit 
is connected to the frame of the machine; the other termi- 
nal C, of the circuit is connected through the voltmeter V 
to the terminal /, of the field coil. If that coil is grounded, 
the voltmeter will show a deflection about equal to the 
E. M. F. of the circuit CC„ but if the insulation is intact, 
it will show little or no deflection. The wire connecting the 
voltmeter with the terminal t^ may be connected in suc- 
cession to the terminal of the other coil, or coils, and to the 
commutator; any grounded coil of the field or armature 
winding will be shown by a considerable deflection of the 
voltmeter needle. 



144. If the machine tests out clear of grounds, it should 
be shut down after the proper length of time and the 



Digitized by VjOOQIC 



§15 



DIRECT-CURRENT MOTORS 



97 



various parts of the machine felt over to locate any excess- 
ive heating. If accurate results are wanted, thermometers 
should be used, placing the bulb on the various parts (arma- 
ture, field coils, etc.) and covering it with a wad of waste 
or rags. They should be looked at from time to time until 
it is seen that the mercury no longer rises, when the point 
to which it has risen should be noted. 



146. Test for Defects in Armatiire (Bar -to -Bar 

Test). — Faults in armatures may best be located by what is 
known as a bar-to-bar test. This consists briefly in send- 
ing a current through 
the armature (in at one 
side of the commutator 
and out at the opposite 
side) and measuring the 
drop between adjacent 
bars all around the com- 
mutator. If the arma- 
ture has no faults, the 
drop from bar to bar 
should be the same for 
all the bars. The con- 
nections for this test are 
shown in Fig. 59. E is 
the line from which the 
current for testing is 
obtained, and L,B, a, 
lamp bank by means of 
which the current flow- 
ing through the arma- 
ture may be adjusted. 
Connection is made with 
the commutator at two 
opposite points ^, ^. A Pio. » 

contact piece, or crab C, is provided with two spring contacts 
that are spaced so as to rest on adjacent bars. These con- 
tacts are connected to a galvanometer, or millivoltmeter, G, 




Digitized by VjOOQ IC 



98 DIRECT-CURRENT MOTORS §16 

For the sake of illustrating the way in which the bar-to-bar 
test will indicate various kinds of faults, we will suppose 
that in coil A^ there is a short circuit, that the commutator 
leads of coils 5, K^ and W have been mixed, as shown, 
and that there is an open circuit in coil T. Current will 
flow through the top coils from A to jff, but not through 
the bottom coils on account of the open circuit at T. Ter- 
minals A^ B may be clamped permanently in place by means 
of a wooden clamp. 

146. It is evident that the deflection of the galvanom- 
eter will depend on the difference of potential between 
the bars. If everything is all right, practically the same 
deflection will be obtained all around the commutator, no 
matter on what pair of bars C may rest. The test is 
carried out as follows: Adjust the lamp bank until the 
galvanometer, or voltmeter, gives a good readable deflection 
when C is in contact with what are supposed to be good 
coils. The amount of current required in the main circuit 
will depend on the resistance of the armature under test. 
If the armature is of high resistance, a comparatively small 
current will give sufficient drop between the bars; if of low 
resistance, a large current will be necessary. The operator 
runs over several bars and gets what is called the standard 
deflection and then compares all the other deflections with 
this. In case he should start on the damaged part, he will 
find when he comes to the good coils a difference in 
deflection. 

If the contact rests on bars 5, 4, it is easily seen that a 
deflection much larger (about double) than the standard 
will be obtained, because two coils are connected between 
S and Jf in place of only one. When on Jf. and 5, the deflection 
of the voltmeter, or galvanometer, would reverse, because 
the leads from K^ 5, and ^Fare crossed. The deflection would 
not be greater than the standard, because only one coil is con- 
nected between -J and 5. Between 5 and 6 a large deflec- 
tion will be obtained for the same reason that a large one 
was obtained between S and J^, Between 6 and 7 little or 



Digitized by VjOOQ IC 



§15 



DIRECT-CURRENT MOTORS 



99 



no deflection will be obtained, because coil 7 is here repre- 
sented as being short-circuited, and hence there will be 
little or no drop through it. As C is moved around on the 
lower side, no deflection will be obtained until bars 15 and 
16 are bridged. There will then be a violent throw of the 
needle, because the voltmeter will be connected to A and B 
through the intervening coils. When C moves on to 16 
and i7, there will again be no deflection, thus locating the 
break in coil T, As a temporary remedy for this, bars 15 
and 16 may be connected together by a jumper ^ or piece of 
short wire. 

147. If any of the coils have poor or loose connections 
with the commutator bars, the effect will be the same as if 
the coil had a higher resistance than it should, and hence 
the galvanometer deflection will be above the normal. In 
practice, after one has become used to this test, faults may 
be located easily and rapidly. It is best to have two per- 
sons, one to move C and the other to watch the deflections 
of G. 

148. liocatingr Short-circuited Armature Coils. 

Where there are a large number of armatures to be tested, 




Fig. 60 



as, for example, in street-railway repair shops, an arrange- 
ment similar to that shown in Fig. 60 is very convenient 



Digitized by VjOOQIC 



100 DIRECT-CURRENT MOTORS §16 

for locating short-circuited coils. ^ is a laminated iron 
core with the polar faces b^b (in this case arranged for four- 
pole armatures). This core is wound with a coil c that i$ 
connected to a source of alternating current. The core is 
built up to a length dy about the same as the length of the 
armature core. The core A is lowered on to the armature, 
and when an alternating current is sent through c, an alter- 
nating magnetization is set up through the armature coils. 
This induces an E. M. F. in each coil ; and if any short cir- 
cuits exist, such heavy local currents are set up that the 
short-circuited coils soon become hot or burn out, thus indi- 
cating their location. If an armature with a short-circuited 
coil is revolved in its own excited field, the faulty coil 
promptly burns out, so that this constitutes another method 
of testing for such faults. To cut out a short-circuited 
coil, temporarily disconnect its ends from the commutator, 
bend back the ends out of the way, tape them so that they 
cannot touch each other, and put a short piece of wire, or 
jumper, in place of the coil so disconnected. It is always 
best, however, to replace the defective coil, because if the 
turns are short-circuited on each other, the coil may persist 
in heating and thus damage other coils. 



Digitized by VjOOQ IC 



ALTERNATING CURRENTS 

(PART 1) 



E. M. F. WAVE FORMS 

1. In studying the applications of electricity, there are, in 
general, two distinct classes of electric currents to deal with, 
namely, direct currents and alternating currents. Most 
of the practical applications of electricity were formerly 
carried out by means of the direct current; but during 
recent years the alternating current has come extensively 
into use. 

2. The apparatus used in connection with alternating- 
current installations is, in general, different from that used 
in connection with direct-current outfits, and must be con- 
sidered separately. Moreover, on account of the nature 
of alternating currents, they do not flow in accordance with 
the simple laws that govern the flow of direct currents. 

3. In continuous-current circuits, the current flows uni- 
formly in one direction ; in other words, as time elapses the 
value of the current does not change. This condition might 
be represented graphically, as shown in Fig. ]. Time is 
measured along the horizontal line a^ and as the current 
remains at the same value, it might be represented by the 
heavy liner^; the height of this line above the horizontal 
would indicate the value of tiie current, i. e., -f- 25 amperes. 
A current of —35 amperes, which would be flowing in the 

§16 

For notice of copyright, see page immediately following the title page. 



Digitized by 



Google 



ALTERNATING CURRENTS 



opposite direction, would be 
line de below the horizontal. 



§16 

represented by the heavy 



4- 






Id 



Y 



\9\ SO M TO 9D 90 109 if 9 190 

Q, Time in Seconds "^^^ 

I 



130 149 160 i99 



PlO.1 



4. In the case of alternating currents, the direction of 
the flow is continually changing. This may also be shown 
graphically, as in Fig. 2. In this case, a current of 25 amperes 
flows for an interval of 1 second in the positive direction, 
then reverses, flows for a similar interval in the opposite 



99 

I" 



I 

^19 

I" 

^4# 



• — i See. 



] 



r 



-1 See,--* 






Time 



PIO 9 

direction, and then reverses again. This operation is 
repeated at regular intervals as shown by the line, and any 
current that passes repeatedly through a set of values in 
equal intervals of time, such as that shown above, is known 
as an alternating: current. The line Off a gg' b is often 



Digitized by VjOOQ IC 



§16 



ALTERNATING CURRENTS 



3 



spoken of as a current, or E. M. F., wave, depending on 
whether the diagram is used to represent the current flowing 
in a circuit or the E. M. F. that is setting up the current. 
The positive half wave Off'a is of almost exactly the same 
shape as the negative half wave agg' b in most practical 
cases. Induction coils produce E. M. F.*s that have differ- 
ent positive and negative half waves, but in the case of 
E. M. F/s produced by alternating-current dynamos the 
two waves are almost identical. 

5. The outline of the alternating-current waves usually 
met with in practice is always more or less irregular, the 




PlO. 8 



shape of the wave depending largely on the construction of 
the alternator producing it. Some of the more common 
shapes met with are shown in Figs. 3, 4, 6, and 6. Figs. 3, 




PIO. 4 



4, and 5 show the general shape of the waves produced by 
some alternators used largely for lighting work and having 
toothed armatures. The student should notice that while 



Digitized by VjOOQIC 



ALTERNATING CURRENTS 



16 



the waves are irregular, the same set of values are repeated 
over and over, and that the set of negative values of the 




Pio. 6 



current is the same as the positive, thus producing a sym- 
metrical curve with reference to the horizontal line. Fig. 6 
represents a form of wave that is commonly met with, 




Pig. e 



especially in the case of large alternators designed for 
power transmission. It will be noticed that this curve is 
practically symmetrical as regards both the horizontal 
line Oabc and a vertical passing through the highest point 
of the curve. 

CTCLE, FREQUENCY, AI.TERNATION, PERIOD 

6. In all the curves shown in Figs. 3 to 6, the current 
passes through a set of positive values, while the interval of 
time, represented by the distance a^ is elapsing, and 
through a similar negative set during the interval repre- 
sented by the distance a b. This operation of passing 



Digitized by VjOOQ IC 



§ 16 ALTERNATING CURRENTS 6 

through a complete set of positive and negative values is 
repeated over and over in equal intervals of time. 

The complete set of values that an alternating current 
passes through repeatedly as time elapses is called a cycle. 
A cycle would therefore be represented by the set of values 
that the current passes through while the time represented 
by the distance b elapses. 

7. The number of cycles passed through in 1 second is 
called the frequency of the current. For example, if the 
current had a frequency of 30, it would mean that it passed 
through 30 complete cycles or sets of values per second. In 
this case the distance Ob would, therefore, represent an 
interval of ^ second, and the time occupied for each half 
wave, or the distance Oa^ would be -^ second. The fre- 
quency is usually denoted by the letter ;/, although the 
letter/" is sometimes used and the symbol ^^ was adopted 
in some of the earlier works on alternating currents^ Fre- 
quencies employed in alternating-current work vary greatly 
and depend largely on the use to which the current is to be 
put. For lighting work, frequencies from 60 to 125 or 130 
are in common use. For power-transmission purposes, the 
frequencies are usually lower, varying from 60 down to 25, 
or even less. Very low frequencies cannot be used for light- 
ing work, because of the flickering of the lamps. Several 
of the large companies have adopted 60 as a standard fre- 
quency for both lighting and power apparatus. This is well 
suited for operating both lights and motors, and enables 
both to be run from the same machme — a considerable 
advantage, especially in small stations. The high frequen- 
cies of 125 to 130 are going out of use except in stations 
that operate lights exclusively. 

8. An alternation is half a cycle. An alternation is, 
therefore, represented by one of the half waves, and there 
are two alternations for every cycle. 

Instead of expressing the frequency of an alternator as so 
many cycles per second, some prefer to give it in terms of so 
many alternations per minute. For example, suppose we 



Digitized by VjOOQ IC 



6 ALTERNATING CURRENTS §16 

have an alternator supplying current at a frequency of 60, 
1. e., 60 complete cycles per second, or 3,600 per minute. 
Since there are two alternations for every cycle, the machine 
might be said to give 7,200 alternations per minute. The 
method of expressing the frequency as so many cycles per 
second is, however, the one most commonly used. 

9. The time of duration of one cycle is called its period. 
This is usually denoted by t and expresses the number of 
seconds or fraction of a second that it takes for one cycle to 
elapse. If the frequency were 60, the period would be 
•^ second, or, in general, 

frequency = :— ? 

^ ^ period 

or n=\ (1) 

10. Two alternating currents are said to be In synclipo- 

nism when they have the same frequency. Two alterna- 
tors would be said, to be running in synchronism when each 
of them was delivering a current that passed through 
exactly the same number of cycles per second. Although, 
strictly speaking, two E. M. F.*s are in synchronism when 
they have the same frequency, the term in synchronism as 
generally used carries with it the idea that not only do the 
two E. M. F.'s have the same frequency, but also that they 
are in phase, i. e., they are in step or pass through their 
maximum and minimum values simultaneously as explained 
later. 

SINE CURVES 

11. The variation of an alternating current, as time 
elapses, may always be represented by a wave-like curve, 
such as those shown in the previous diagrams, such curves 
being easily obtained from the alternators by several well- 
known methods. The current at any instant may thus be 
obtairied. In order, however, to study the effects of an 
alternating current, it is necessary to know the law accord- 
ing to which this curve varies. In the case of the irregular 



Digitized by VjOOQ IC 



§16 



ALTERNATING CURRENTS 



curves shown in Figs. 3, 4, and 5, the law giving the rela- 
tion between the time and the value of the current is so 
complicated that it renders calculations too involved. A 
great many alternators give E. M. F. and current curves 
that closely resemble that shown in Fig. 6, and the law 
that this curve follows is quite simple. Such a curve may 
be constructed as follows : Suppose a point /, Fig. 7, moves 




PIO. 7 

uniformly around a circle in the direction of the arrow, 
starting from 0°. The angle a (pronounced alpha) will 
uniformly increase from 0° to 180° and from there to 360°, 
or back to 0°, and so on. Take the instant when the point 
is in the position shown and project it on the vertical 
through O. The line Of = r/, which is the projection 
of O p^ will be proportional to the sine of the angle or, 



because sin 



a = ^ and 



Op remains constant. When 



a = 90°, the projection, or sin o', is proportional to O h\ 
when at 180° or 360° it is zero, and when at 270° it is pro- 
portional to O //'. All the values of the projection of the 
line Op from 0°to 180° are positive or above the horizontal, 
and those from 180° to 360° are negative or below the hori- 
zontal. As the point / revolves, p' moves from O to A, 
back through O to //', and then back to O when the point/ 
has reached 0° again. The way ir^ which the sine (or length 
of the line Op') varies as the point / revolves may be shown 
by laying out along the line Ox distances representing the 
time that it takes the point p to turn through various values 
of the angle a and erecting perpendiculars equal to the values 
of the sine corresponding to these angles. At the point a 



Digitized by VjOOQ IC 



8 ALTERNATING CURRENTS §16 

the value of the projection of Op would be zero. At the 
instant shown in the figure the value of the sine is r p = O p\ 
Lay off the distance af^ representing the time required for 
the point to move from 0*^ to/, and erect a perpendicular/^ 
equal to r p. The distance a b represents the time required 
for / to move through 90®, and the perpendicular ^^ is 
equal to (9A = Op, A number of points may be found in 
this way and the half wave akc drawn in. The negative 
half wave is exactly the same shape, but, of course, is drawn 
on the lower side of the horizontal. A curve constructed in 
this manner is known as a sine curve, because its perpen- 
dicular at any point is proportional to the sine of the angle 
corresponding to that point. The sine of the angles O'', 
180°, 360° is zero; hence, at the^ points corresponding to 
these angles, the curve cuts the horizontal, i. e., the curve 
passes through its zero value. At 90° the curve passes* 
through its positive maximum value, and at 270° it passes 
through its negative maximum. The maximum value of 
the curve bk ■=^ Op is called its amplitude, and the curve 
varies between the limits -{-b k and —bk, 

13. If an alternating current or E. M. F. is represented 
by a sine curve constructed as above, the maximum value 
•that the current or E. M. F. reaches during a cycle will be 
represented to scale by the vertical bk^ and the value at 
any other instant during the cycle will h^ b k sin cr, where a 
is the angle corresponding to the instant under considera- 
tion. The law that such a curve follows is therefore quite 
simple, and fortunately such a curve represents quite closely 
the E. M. F. and current waves generated by a large class 
of alternators. Even where the curves do hot follow the 
sine law exactly, it is sufficiently accurate for most prac- 
tical purposes to assume that they do. In calculations con- 
nected with alternating currents, it is therefore usual to 
assume that the sine law holds good. The wave shown in 
Fig. 7 may therefore be taken to represent the way in 
which an alternating current varies. During the time rep- 
$ented by the distance a e^ one complete cycle occurs. The 



Digitized by VjOOQ IC 



§16 ALTERNATING CURRENTS 9 

maximum value of the current is represented by the ver- 
tical, or ordinate, \ = d k, and the current passes through 
a certain number of these cycles every second, depending 
on the frequency n of the alternator. Since the revolving 
line Op = bk = \^ it follows that Op represents ihe maxi- 
mum value of the current, and its projection on Oh at any 
instant represents the value of the current at that instant. 

13. The student should always keep in mind the fact 
that in an alternating-current circuit there is a continual 
surging back and forth of current, rather than a steady flow, 
as in the case of direct current, and it is this continual 
changing of the current that gives rise to most of the 
peculiarities that distinguish the action of alternating cur- 
rents from direct. 

PROPERTIES OF SINE CITRVES 

14. By examining Fig. 7, it will be seen that every time 
the point p makes one complete revolution, the sine curve 
passes through one complete cycle, and the time that it 
takes p to make one revolution corresponds to the period. 
The number of revolutions that the point / makes in 
1 second would therefore be equal to the frequency ;/. It is 
well to note, in passing, that the sine curve is much steeper 
when it crosses the axis than when it is near its maximum 
values; in other words, the sine of the angle is changing 
more rapidly when near 0° and 180° than when near 90° 
and 270°. 

ADDITION OF SINE CURVES 

15. If a line carrying a continuous current is split into 
two or more branches, the current flowing in the main line 
is found by taking the sum of the currents in the different 
branches. For example, suppose a pair of electric-light 
mains feeds three circuits, taking 5, 10, and 50 amperes; 
the current in the mains would be found by simply taking 
the sum of these, i. e., G5 amperes. This method, however, 
cannot^ as a general rule, be applied to alternating-current 



Digitized by VjOOQ IC 



10 



ALTERNATING CURRENTS 



§16 



circuits, and it is necessary, therefore, to study carefully 
the methods of adding together two or more alternating 
currents or E. M. F. 's. 

16. The method of adding together two alternating 
E. M. F.'s represented by sine curves is shown in Fig. 8. 




PlO. 8 

One E. M. F. is represented to scale by the radius Op\ that 
is, the radius Op is laid off to represent the maximum value E 
that the E. M. F. reaches during a cycle. The point / is 
supposed to revolve uniformly around the inner circle, and 
the corresponding sine curve is shown by the dotted wave. 
The other E. M. F. has its maximum value E' represented 
by the radius O q^ which revolves uniformly around the 
outer circle, generating the sine wave shown by the dot-and- 
dash line. Both points are supposed to revolve at exactly 
the same speed and to start from 0° at the same instant; in 
other words, the frequency of both is the same, or they are 
in synchronism. Since both points start from 0*^ at the 
same instant and revolve at the same rate, it follows that 
both the E. M. F. curves will vary together. They will 
come to their maximum values at the same instant and will 
pass through zero simultaneously. When two or more alter- 
nating E. M. F.'s or currents vary together in this way, 
they are said to be In phase with each other. The curve 



Digitized by VjOOQ IC 



§16 ALTERNATING CURRENTS 11 

representing the sum of these two E. M. F.*s is easily found 
by adding together the instantaneous values of the separate 
E. M. F.'s, thus giving the sine curve shown by the full line. 
For example, take any instant represented by the point /on 
the line Ox; the ordinate, or vertical, fg^ represents the 
instantaneous value of the E. M. F. E, and the ordinate /A 
represents the instantaneous value of E'. The value of their 
sum at this particular instant is therefore found by adding 
/ £■ and ///, giving / k, and locating the point k on the 
required curve. The maximum ordinate of this resultant 
curve is ^/ = E + E'. This curve representing the sum of 
the two original E. M. F.'s is in phase with E and E', and is 
also a sine curve; hence, it may also be represented by a 
line revolving about the point (?, provided this line is so 
taken that its projection on the vertical is at all instants 
equal to the sum of the projections of the two original lines 
0/f and Og. Since the points/ and q are in line with each 
other, it follows that if we produce O g to r, making Or 
= Op + Oq, the projection of Or = Or' will be equal to 
the sum of the projections of Op and O q, i, ^.^ O r' = Op' 
+ Oq'. It follows, then, from the above, that if the line Or 
were to revolve uniformly around O at the same rate as 
Op and Oq, the curve representing the sum of the two 
E. M. F.'s would be generated. 

17. The following may be summarized from the fore- 
going: 

1. The curve representing the sum of two or more sine 
curves may be obtained by adding together the ordinates repre- 
senting the instantaneous values of the original curves, 

2. If the two or more curves that are added are of the 
same frequency {as is usually the case), the resultant also will 
be a sine curve. 

2. Two or more alternating E. M, F. *s or currents of the 
same frequency are said to be in phase with each other, when 
they reach their maximum and minimum values at the same 
instant. 

44—24 



Digitized by VjOOQ IC 



12 ALTERNATING CURRENTS §16 

4. The resultant sine curve representing the sunt of two or 
mare sine curves of the same frequency may be generated by a 
line revolving uniformly^ and of such length and so located 
with reference to the lines generating the original curves that 
its projection on the vertical will at all instants be equal to the 
sum of the projections of the two original revolving lines. 

18. In Fig. 8, the line Or = Op + O q generates the 
resultant curve, and since Op and Oq are in phase with 
each other, it is seen that (9r is found by simply taking 
the sum of the two original lines Op and O q. In other 
words, when two alternating E. M. F/s or currents are 
in phase, they may be added together in the same way 
as direct currents, but if they are not in phase, this can- 
not be done. It is quite possible in alternating-current 
circuits to have the current and E. M. F. out of phase with 
each other, and the same is true regarding two or more 
E. M. F. *s and two or more currents. Suppose that an alter- 
nator is forcing current through a circuit. Every wave of 
E. M. F. will be accompanied by a corresponding wave of 
current, and hence the current and E. M. F. will always 
have the same frequency. There are a number of causes 
that may prevent the current and E. M. F. from coming to 
their maximum and minimum values at the same instant, 
and the current may lag behind the E. M. F. or may be 
ahead of it. In the case of two alternators feeding current 
into a common circuit, one current may lag behind the 
other, or the E. M. F. produced by an alternator may not 
be in phase with its current. The effects of this difference 
of phase must be taken into account in alternating-current 
work, as it gives rise to a number of peculiar effects not met 
with in connection with direct currents. Of course, in the 
case of direct currents there is no change taking place in 
the currents or E. M. F.'s; consequently, in direct-current 
circuits, there cannot be any phase difference between cur- 
rent and E. M. F. 

19. The effects of this difference in phase will be seen 
more clearly by referring to Fig. 9. The two E. M. F.*s E 



Digitized by VjOOQ IC 



§ 16 ALTERNATING CURRENTS 13 

and E' are represented by the revolving lines Op and Oq as 
before. In this case, however, the E. M. F. E' starts from 
0° before E, and the two are displaced by the angle (phi) ; 
in other words, the point / does not start from 0"^ until the 
point q has turned through an angle 0. After this, the two 
revolve in synchronism, and the angle remains constant, 
while the angle a is continually changing, as before. The 
E. M. F. E is therefore lagging behind E', and the two 
E. M. F. 's are said to be out of phase. It might also be 
said that the E. M. F. E' was ^ degrees ahead or in advance 
of E, or that E lagged ^ degrees behind E'. If E lagged 
one-half a cycle behind E', the angle of lag would be 180°| 



PfO. 9 

and if it were said that E lagged 90** behind E', it would 
mean that E came to its maximum or minimum value just 
one-fourth of a period later than E'. The dot-and-dash 
curve, as before, represents E', and the dotted curve repre- 
sents E. The two curves, however, no longer cross the 
horizontal, or come to their maximum, at the same instant. 
The dotted curve starts in at the point r, a distance ac 
behind the curve representing E'. Since time is measured 
in the direction of the arrow, the point c represents an 
instant that is later than the point a. In other words, the 
curve representing E does not start until an interval of 
time, represented by ac^ has elapsed after the starting of 
the dot-and-dash curve, i. e., the dotted curve is lagging 



Digitized by VjOOQ IC 



14 ALTERNATING CURRENTS §16 

behind the other. The distance acis equivalent to the time 
that it would take / or ^ to turn through the angle 0. The 
sum of the two curves is found, as before, by adding the 
ordinates; thus hk -\- hi •= h m^ and the resultant curve 
shown by the full line is obtained. It will be noticed that 
this is a sine curve, but it is in phase with neither of the 
original curves, also that its maximum value E " is not as 
great as in the case shown in Fig. 8. The resultant curve 
has, however, the same frequency as the others. The 
revolving line that will generate this curve must have its 
projections on the vertical equal to the sum of the projec- 
tions of the other two. This condition is fulfilled by the 
line O r, which is the diagonal of the parallelogram formed 
on the two sides Op and O q. That this is the case will be 
seen by referring to the figure. The projection of Op is Op' 
and of O q, O q\ The projection of the diagonal Or\^Or'\ 
but q' r' = (9/; hence. Or' = O q' + O p\ and if the diag- 
onal Or were to revolve at the same rate as Op and O q^ the 
full-line curve would be generated. The maximum value of 
this curve E" is equal to (?r, and the diagonal Or not only 
gives the maximum value of the resultant curve, but also 
gives its phase relation in regard to the two original curves. 
In this case Or lags behind qhy the angle /? (beta), and 
the distance a b represents the interval of time that is 
required for the line O r \,o swing through the angle /?. 

20. The important points in the preceding may then be 
summarized as follows: 

1. In alternating-current systems, two or more currents 
or E, M. F/s may not come to their maximum and minimum 
values at the same instant, in which case they are said to be 
out of phase or to differ in phase, 

2. Phase difference is usually expressed by an angle (in 
most cases denoted by the Greek letter <t>). If this angle is 
measured forwards, in the direction of rotation, it is called an 
an^le of lead, or ao^le of advance; if ^Pleasured backwards^ 
it is called an angrle of lag:. 



Digitized by VjOOQ IC 



§16 



ALTERNATING CURRENTS 



16 



3. Two sine curves not in phase may be added together 
by adding their ordinates. If the two original curves are of 
the same frequency^ the resultant curve will be also a sine 
curve ^ but differing in phase from the other two, 

4. The maximum value of the resultant curve is given 
by the diagonal of the parallelogram constructed on the lifies 
representing the maximum values of the original curves. 
This diagonal not only gives the maximum value of the 
resultant curve ^ but also determines its phase relation. 

21. From the foregoing it will be seen at once that in 
adding alternating currents and E. M. F.*s, account must 
be taken not only of their magnitude, but also of their phase 
relation, and they can be added numerically only when they 
are in phase, as in the case of direct currents. 

22. As an example of the addition of two alternating 
currents, suppose a circuit is divided as shown in Fig. 10. 




PIO. 10 

The main circuit 1 is divided into the branches 2 and 5, and 
ammeters -^„ A^^ and A^ are placed in the branches. If 
a continuous current were flowing, the reading of A^ would 
be equal to the sum of the readings of A^ and A^, This, 
however, would not be the case if an alternating current 
were flowing, unless the currents in the two branches hap- 
pened to be exactly in phase. Generally speaking, they 
would not be in phase, and the reading of A^ would be less 



Digitized by VjOOQ IC 



16 



ALTERNATING CURRENTS 



16 



than the sum of A^ and A^, The relation between the three 
currents would be as shown in Fig. 11, where I,, the reading 
of A^, is the diagonal of the parallelogram formed by O g 

and (7/, representing the 
readings of A^ and A^, 
respectively. These lines 
can all be laid off to 
scale, so many amperes 
per inch, and the angles 
of phase difference read- 
ily determined. In this 
case is the phase difference between the current in the 
branches S and ^, while the main current is fi° behind I, and 
6° (theta) ahead of I,. The resultant current represented 
by Or is smaller than it would be ii Og and Op were in 




FIO. 11 



ot 



1 



PIO. 12 



phase. If Op and Oq were in phase, they could be added 
together directly, and the resultant would be Or, as shown 
in Fig. 12. Here the angle of lag has become zero, and 
the parallelogram has reduced to a straight line. 



X^m 




Pig. 18 



33« A practical example of the addition of two alterna- 
ting currents is shown in Fig. 13 where two alternating- 



Digitized by VjOOQIC 



16 



ALTERNATING CURRENTS 



11 



current motors A and B are supplied with current from the 
same line. In this case, the two motors A and B take cur- 
rents that will in all probability be out of phase and the 
actual current flowing in the line will be found by taking 
the geometric sum of the two separate currents, as explained 
in the preceding example. 

34. Alternating E. M. F.'s are added in the same way 
as currents. If it were possible to operate two alter- 
nators A and B^ Fig. 14, in series, one giving an E. M. F. 



f4nf 




Pio. 14 

E, and the other E, volts, the E. M. F. E, obtained across 
the mains would not generally be equal to E, + E„ but 
would be the resultant sum, as shown by the parallelogram, 
Fig. 15. If A and B were continuous-current dynamos, or 
if E, and E, were exactly 
in phase, E, would be 
equal to the sum of E, 
and E,. On account of 
this effect of the differ- 
ence in phase between 
E, and E,, the sum of 
the readings of voltmeters connected across a, a' and ^, b' 
will be greater than the reading obtained by a voltmeter 
connected across the mains. Alternators operated in series 
would have to be rigidly connected, otherwise the phase 
relation between their E. M. F.'s would be continually 
changing. 




PlO. 15 



Digitized by VjOOQ IC 



18 



ALTERNATING CURRENTS 



16 



TWO-PHASE AXD THREE-PHASE STSTEMS 

26« If the angle of lag between two currents is zero, they 
are said to be in phase. 

If the angle of lag between two currents is 90°, they are 
said to be at rlgrlit angrles, or in 
quadrature. 

If the angle of lag is 180**, they 
are said to be in opposition. 

Fig. 16 shows two current waves 
at right angles or in quadrature, the 
current l„ represented by the dotted 
line, lagging 90°, or \ cycle, behind I,. 
If these two currents were fed into 
a common circuit, the resultant cur- 
rent would be represented by the 
diagonal <?r, and this current would 
lag 45° behind 1^. 

36. It will be seen by examining 
Fig. 16 that at the instant Ij is at its 
maximum value, I, is passing through 
zero. If each of these currents were 
fed into separate lines, a t^vo-phase, 
or quarter-phase, system would be 
obtained, i. e., there would be two 
distinct circuits, fed from one dyna- 
mo, the currents in the two circuits 
differing in phase by 90°, or one- 
fourth of a period. Such systems 
are in common use for operating 
motors, and will be more fully ex- 
plained in connection with alterna- 
tors. Alternators for use with such 
systems are usually provided with 
two sets of windings on their armatures, so arranged that 
when one set is generating its maximum E. M. F., the other 
is passing through zero. 




Digitized by VjOOQIC 



§16 



ALTERNATING CURRENTS 



Id 



27. Systems in which three currents are employed are 
also in common use in connection with power-transmis- 
sion plants. These currents differ in phase by 120**, or 
one-third of a cycle, and constitute 
what is known as a three-phase 
system. Such an arrangement is 
shown in Fig. 17, where the three equal 
currents l„ l„ and I, are displaced in 
phase by 120°. The corresponding 
sine curves are also shown, the dotted 
line I, lagging 120° behind the full 
line l„ and the dot-and-dash line 1. 
f' ^s I lagging 120° behind I,. In such a case, 

/ x^y^ where the three currents are equal, 

I /\ the resultant sum is at all instants 

\ y*^ \ equal to zero. This may be seen from 

the curves, gl^ for example, being the 
equal and opposite of gh-^-gk^ and 
2 de the equal and opposite of df. It 
is well to bear this property of a bal- 
anced three-phase system in mind, as 
it is taken advantage of in connection 
with three-phase armatures and in 
three-phase power-transmission lines. 
The above would not be true if the 
system were unbalanced, i. e., if I,, l„ 
and I, were not all equal ; but in most 
cases where these systems are used, it 
is tried as far as possible to keep the 
currents in the different lines equal. 
It should also be noted that in such a 
system when the current in one line 
is zero, the currents in the other two 
lines are equal and are flowing in 
opposite directions. When the cur- 
rent in one circuit is at its maximum value, the currents 
in the other two are in the opposite direction and one- 
half as great. 




Digitized by VjOOQ IC 



ALTERNATING CURRENTS 



16 



COMPOSITION 



AND RE80I.UTI0N 
AND E. M. F.'S 



OF . CURRENTS 



Lim 




Pio. 18 



28. It has been shown that a sine wave may always be 
represented by a line revolving uniformly around a point 

and showing to scale 
the maximum value 
of the E. M. F. or 
current. The pro- 
jection of this line 
on the vertical at 
any instant gives 
the instantaneous 
value of the E. M. F. 
or current. In working out problems in connection with 
alternating currents, it is not necessary to draw in the sine 
curves, but to use simply the line representing the curve. 
It has already been shown how currents and E. M. F. *s may 
be added by using these lines. In fact, alternating currents 
and E. M. F.'s are added and resolved into components in 
just the same way as forces are treated in mechanics, by 
means of the parallelogram of forces. What holds true with 
regard to two currents also applies to three or more, in this 
case the polygon of forces being employed. An example of 
this is shown in'Fig. 18. Three alternating-current motors 
A^ By and C are in parallel, and supplied with current 
from the same line. The three currents differ in phase, and 
their amounts are given by the amme- 
ters l„ I,, I,. Required, the current flow- 
ing in the main circuit. This will be the 
resultant sum of I,, l„ and I,. Lay off O a^ 
Ob, and Oc, Fig. 19, to represent the 
three currents to scale and in their proper 
phase relation. From a draw a ^ equal 
and parallel to Ob, and from e draw e d^ 
equal and parallel to O c. Join O d ; then 
Od will represent the resultant current 
to the same scale that O a, Ob, and Oc represent l„ l„ 




FlO. 19 



Digitized by VjOOQ IC 



§16 ALTERNATING CURRENTS 21 

and I,. In other words, Od will represent to scale the 
reading of the ammeter I placed in the main line. The 
angles giving the phase differences between the different 
currents can usually be calculated when the constants of the 
different circuits are known. Methods for calculating the 
angles of phase difference will be given later. If the three 
currents represented \>y O a^ O b^ and O c should all happen 
to be in phase with one another, their resultant sum can be 
obtained by simply adding them up numerically. 

29. Alternating E. M. F.*s and currents maybe resolved 
into two or more components in the same way that forces 
are resolved in mechan- - 
ics, by reversing the T 
process of composition. 
For example, in Fig. 20, 
we have the current 
represented by Oa re- 
solved into two com- 
ponents O b and O c. The 
two currents represented 
hy Ob and Oc^ differ- 
ing in phase by the „ _ 
angle 0, will therefore 
combine to produce the current represented by O a. 




EXAMPLES FOR PRACTICE 

1. Two currents, one of 40 amperes and the other of 50 amperes, 
differing in phase by 30°, unite to form a third current. Required, the 
value of this current. Ans. 87 amperes, nearly 

2. Construct a curve that will represent the E. M. F. of an alter- 
nator generating a maximum of 300 volts at a frequency of W. Use 
1 inch per 100 volts for the vertical scale and 1 inch equal to jj^ second 
for the horizontal. 

3. Represent two E. M. F.'s of same frequency, one of 200 volts 
maximum and the other of 150 volts maximum, the latter lagging 
behind the former by an angle of 60**. Draw the two sine curves 



Digitized by VjOOQ IC 



22 ALTERNATING CURRENTS §16 

representing these E. M. F. 's in their proper relation to each other, and 
add these together to obtain the resultant curve. Find the maximum 
value of this resultant and compare its value with that obtained by 
taking the diagonal of the parallelogram constructed from the two 
component E. M. F.'s. Ans. Resultant E. M. F. = 804 volts 

Note. — The student should bear in mind that in all the foregoing 
cases the waves of current and E. M. F. are supposed to follow the 
sine law. 



MAXIMUM, AVERAGE, AND EFFECTTTB VAIiUBS 
OP SINE WAVES 

30. During each cycle, an alternating current passes 
through a large range of values from zero to its maximum. 
These instantaneous values are, as a rule, used very little in 
calculations. It is necessary to have it clearly understood 
what is meant when it is said that a current of so many 
amperes is flowing in a circuit or that an alternator is sup- 
plying a pressure of so many volts. When it is stated that 
an alternating current of, say, 10 amperes is flowing in a 
circuit, some average value must be implied, because, as a 
matter of fact, the current is continually alternating through 
a range of values. It has become the universal custom to 
express alternating currents in terms of the value of the 
continuous current that would produce the same heating 
effect in the circuit; as, for example, if the alternating cur- 
rent were 10 amperes, it would mean that this alternating 
current would produce the same heating effect as 10 amperes 
continuous current. 

31. Suppose the sine curve. Fig. 21, represents the 
variation of an alternating E. M. F. ; there are three values 
that are of particular importance: 

1. The maximum Talue, or the highest value that the 
E. M. F. reaches, is given by the ordinate E. This maxi- 
mum value is not used to any great extent, but it shows the 
maximum to which the E. M. F. rises, and hence would 
indicate the maximum strain to which the insulation of the 
alternator would be subjected. 



Digitized by VjOOQ IC 



§16 



ALTERNATING CURRENTS 



23 



2. The average value of a sine curve is the average of 
all the ordinates of the curve for one-half a cycle. For 
example, in Fig. 21, the average ordinate of the curve a/ d 




PlO. 21 



would be that ordinate a d which, multiplied by the base^^, 
would give a rectangle abedoi the same area as the sur- 
face afb. The average value taken for a whole cycle would 
be zero, because the average ordinate for the negative wave 
would be the equal and opposite of the positive ordinate. In 
the case of a sine curve, this average value always bears a 
definite relation to the maximum value.* If E is the maxi- 

2 E 

mum value, the average value is — , or .636 E.* The aver- 

age value is used in some calculations, but, like the maximum 
value, its use is not very extended. The relation between 
the average and maximum value, however, is used consider- 
ably and should be kept in mind. 

3. The effective value of an alternating current may be 
defined as that value which would produce the same heating 



© 



2 

♦The fact that the average value of the sine is .686, or-, may be 

proved exactly by means of calculus, but the student can easily verify 
It approximately by taking the values of the sine of the angles from 
0' to 90° from a trigonometric table, adding these together, and divi- 
ding by the number of values. The average so found will be very 
nearly .686. The value can be obtained with a fair degree of accuracy 
by adding the sines of every fifth degree, as, for example, 0**, 5^, 10 , 
15% etc. 



Digitized by VjOOQ IC 



24 ALTERNATING CURRENTS §16 

effect in a circuit as a continuous current of the same 
amount. This effective value is the one universally used to 
express alternating currents and E. M. F.'s. It always 
bears ^a definite relation to the maximum value. When 
ammeters or voltmeters are connected in alternating cir- 
cuits, they always read effective amperes or volts, though, 
for purposes of illustration, we have assumed in some 
of the previous examples connected with the composi- 
tion and resolution of E. M. F.'s and currents that 
maximum values are indicated. This effective value is 
not the same as the average value (.636 maximum), 
as might at first be supposed, but it is slightly greater, 
being equal to .707 times the maximum value. If a contin- 
uous current / be sent through a wire of resistance R^ the 
wire becomes heated, and the power expended in heating 
the wire is IV = P R watts, or is proportional to the square 
of the current. If an alternating current be sent through 
the same wire, the heating effect is at each instant propor- 
tional to the square of the current at that instant. The 
average heating effect will therefore be proportional to the 
average of the squares of all the different instantaneous 
values of the current, and the effective value of the current 
will be the square root of the average of the squares of the 
instantaneous values. The effective value is for this reason 
sometimes called the square-root-of-mean-square value. It 
is also frequently called the virtual value. Suppose, for 
example, a circuit in which an alternating current of 
10 amperes maximum is flowing. This means that the 
current is continually alternating between the limits 
+ 10 amperes and — 10 amperes, and passing through all 
the intermediate values during each cycle. Now, as far 
as the heating effect of this current is concerned, it 
would be just the same as if a steady current of .707 X 10, 
or 7.07 amperes, were flowing, and if an ammeter were 
placed in the circuit, it would indicate 7.07 amperes.* 
Hereafter, in speaking of alternating E. M. F.'s and cur- 
rents, effective values will be understood unless other- 
wise specified. 



Digitized by VjOOQ IC 



§16 ALTERNATING CURRENTS 25 

RELATIONS BETWTEEN VALUES 

33. The relation between the maximum, average, and 
effective values will be seen by referring to Fig. 22, the 




Pig. 22 

average ordinate .636 E being slightly shorter than tTie effec- 
tive .707 E. For convenience, the following relations are 
here given together. They should be kept well in mind, as 
they are used continually in problems connected with alter- 
nating-current work. 

Average value = .636 maximum value 

T:^ce i.- 1 Nr^i^ ' 1 maximum value 
Effective value = .707 maximum value, or — 

Effective value = 1.11 average value 

effective value 

33« Form Factor. — The ratio ; — is known 

average value 

as the form factor of an E. M. F. or current wave, because 

this ratio depends on the shape of the wave. For a sine 

wave, it has just been shown that the form factor is 1.11. 

For a peaked wave, like that shown in Fig. 4, it will be 

greater than 1.11, and for the flat wave shown in Fig. 2, it 

will be equal to 1. 

34. On account of the importance of the effective value, 
the following proof is given of the relation: Effective 



Digitized by VjOOQIC 



26 



ALTERNATING CURRENTS 



§16 



value = .707 maximum. 




or 



Pig. 28 



sin* a -f- cos* a = 



ab^ 



Od"^ Od 



Let O a^ Fig. 23, represent the 
maximum value of the sine. The 
sine at any instant correspond- 
ing to the angle a is a ^ = O a 
sin «, 

ab 
Oa 

Ob 
Oa 
Ob' 
Oa' 
Ob' __ ab'+Ob' Oa' 
Oa' 



sm a = 



sm" a = 



cos a = 



cos" a = 






Oa 



Now as the line Oa revolves, thus generating the sine 
curve, the sine varies from to O a and the cosine varies 
from O c {= O a) to 0^ so that the sine and cosine pass 
through the same range of values, and consequently the 
average of the squares of the sine and the cosine must be 
the same. Since 

sin* a -f- cos* or = 1 

av. sin* a + av. cos* « = 1 

From the above, 2 av. sin* « = 1 

av. sin* or = ^ 

Vav. sm' a = — = 
4/2 

As the alternating E. M. F. is supposed to follow the sine 
law, the instantaneous value of the E. M. F. at any instant 
corresponding to the angle a, is ^ = E sin a^ where E is the 
maximum value; hence, 

e' = E* sin* a 
av. e' = E* av. sin* a 

av. e' = -zr- 



Digitized by VjOOQIC 



§16 



ALTERNATING CURRENTS 



27 



i. e., the effective^ or square-root-of -mean-square value ^ is equal 
to the maximum value multiplied by -T=r, or .707. 

36. Hereafter, in designating E. M. F.'s and currents 
we will use heavy-faced type to denote maximum values of 
E. M. F/s or currents, and light-faced type to denote the 
effective values. In most of the previous figures, the values 
used were maximum values; hence, the heavy-faced letters 
E and I were employed. Instantaneous values will be 
expressed by small italic letters, indicated in the following 
list of symbols: 

E = maximum E. M. F. ; 

E = effective, square-root-of-mean-square, or virtual 

E. M. F. ; 
e = instantaneous E. M. F. ; 
I = maximum current; 
/ = effective, square-root-of-mean-square, or virtual 

current; 
i = instantaneous current. 




PIO. 94 

36. In Fig. 24, the value of the square-root-of-mean- 
square ordinate is obtained graphically. The dotted curve 

44—25 



Digitized by VjOOQIC 



28 ALTERNATING CURRENTS §16 

is obtained by squaring the ordinates of the sine curve, and 
the area of the dotted rectangle is equal to the area enclosed 
by the curve of squares. The height of this ordinate is 

—, and it represents the average of all the values of ^. 

The square root of this gives the effective value of the sine 

curve: —= = .707 E. 
/2 

37. In the preceding diagrams, showing the composition 
and resolution of E. M. F/s and currents by means of the 
^ _ parallelogram, maximum 

X' ^---'^'''"^^^'^ values were used. Since, 

^V^ ^ — ' y however, the effective 

^/SS*^**^^ y or virtual values always 

^ JB, * * bear a fixed proportion to 

P'o- » the maximum, it follows 

that this construction will apply equally well in case effect- 
ive values are used. For example, if the two virtual 
E. M. F.*s E^ and E^ are represented by the lines Ob and O a^ 
Fig. 25, the line Oc will represent the resultant virtual 
E. M. F. to the same scale that Oa and O b represented the 
original quantities. 



SELF-rNDUCTION AND CAPACITT 

38. In most cases the flow of an alternating current 

through a circuit differs considerably from that of a direct 

current, even though the E. M. F. is the same in both cases. 

If a given direct E. M. F. E^ is applied to a circuit of 

resistance R^ the current /' that will flow, determined by 

E' 
Ohm*s law, is /' = ^. If an alternating E. M. F. having an 

effective value E of the same amount as E* be applied to the 
same circuit, the resulting current / may or may not have 
the same value as /'. If there were no self-induction or ca- 
pacity present in the circuit, the alternating current would 
behave in the same way as the direct current, but in most 



Digitized by VjOOQ IC 



§16 ALTERNATING CURRENTS 29 

cases there is more or less self-induction or capacity present 
so that the current will not flow in accordance with the 
simple law that governs direct currents. Most of the devices 
used in connection with the applications of alternating cur- 
rents have more or less self-induction ; alternating-current 
motors, arc lamps, transformers, and even the transmission 
lines, all have it in greater or less amount. Overhead lines of 
considerable length also have an appreciable amount of elec- 
trostatic capacity, and the electrostatic capacity of under- 
ground cables is quite large. Under some conditions, certain 
kinds of alternating-current motors also act as if they had 
electrostatic capacity. Generally speaking, the effects of 
capacity are not as commonly met with as those of self- 
induction, but they are of great importance nevertheless. 

39. Effects of Self-induction. — It has already been 
shown that any circuit or device that is capable of setting up 
a magnetic field through itself possesses self-induction. A 
simple example of this is a coil of wire wound on an iron 
core. When current is sent around the wire, a magnetic 
field is set up through the coil. It was further shown that, 
while the magnetic field is being set up, an E. M. F. is 
induced in the coil, which is opposed to the current. When 
the circuit is broken, the lines of force threading the coil 
collapse, thus inducing an E. M. F. that tends to maintain 
the current; as long as the current flows steadily, there is 
no change in the number of lines of force passing through 
the coil, and, hence, no E. M. F. is induced. This is the 
state of affairs when a direct current is flowing so that, 
except at the moment when the circuit is made or broken, 
the self-induction has no influence on the flow of a direct 
current. When, however, a current that is continually 
changing its value, as, for example, an alternating current, 
is sent through the coil, an induced E. M. F. is present all 
the time that the current is flowing, and this E. M. F. has a 
marked influence on the flow of the current. The induced 
E. M. F. tends to oppose any change in the current, thereby 
choking it back and making the circuit appear as if it had a 



Digitized by VjOOQ IC 



30 ALTERNATING CURRENTS §16 

higher resistance than when a direct current was sent 
through it. In addition to the self-induction increasing the 
apparent resistance of the circuit, it makes it behave as if it 
possessed inertia, and the current does not respond at once 
to the changes in the applied E. M. F. but lags behind it, 
thus throwing the current and E. M. F. out of phase. The 
effects of capacity are exactly the opposite to those of self- 
induction, as will be shown later. 

40. In dealing with alternating-current problems that 
arise in practice, we have the following classes of circuits to 
consider : Circuits containing resistance only, circuits con- 
taining self-induction only, circuits containing resistance 
and self-induction, circuits containing capacity only, circuits 
containing resistance and capacity, circuits containing self- 
induction and capacity, and circuits containing resistance, 
self-induction and capacity. 



CIRCUITS CONTAINING RESISTANCE ONXT 

41. It is possible to have an alternating-current circuit 
or device that possesses no self-induction, or at least where 
the self-induction is so small as to be almost negligible. If 
coils are wound by first doubling the wire on itself, as 
explained in connection with resistance boxes, they have 
practically no self-induction, because the magnetizing action 
of one-half of the turns is neutralized by that of the other 
half; a water rheostat or a load of incandescent lamps has 
very little self-induction. Such devices constitute a non- 
inductive load. 

42. If an alternating E. M. F. E is applied to a non- 
inductive resistance R, the current s^t up through the 

E 
resistance will be I = —^ because there is no induced 

K 

E. M. F. present to choke back the current, and the only 

thing that opposes the flow of the current is the ohmic 

resistance R. By the ohmic resistance is here meant that 

resistance which the circuit offers on account of its inherent 



Digitized by VjOOQ IC 



§16 ALTERNATING CURRENTS 31 

properties as a conductor, i. e. , the resistance that depends 

on the length, cross-section, and material of the conductor. 

Also, since £ = / R, it follows, since the value of R is fixed, 

that when E is at its maximum value, / must also be at its 

maximum value, and when E is passing through zero, / is 

also passing through zero. From the foregoing, then, we 

may state that: 

W/ien an alternating E, M, F. is applied to a non-inductive 

circuity the current flows in accordance with Ohm's law^ 

E 
/ = -=, and^ hence, is of the same amount as if a direct 
K 

E, M. F, of like value were applied to the circuit. 

When an alternating current flows through a non-inductive 

circuit, the current is in phase with the applied E, M. F. 

Example. — An alternating E. M. F. of 300 volts is applied to a load 
of lamps having a combined resistance of 2 ohms, {a) What current 
will flow in the circuit ? {b) What will be the phase difference between 
the current and E. M. F. ? 

Solution. — {a) Since the resistance is non-inductive, the current 
will flow according to Ohm's law and will be *4^ = 150 amperes. Ans. 

{b) The current will be in phase with the E. M. F. ; hence, the 
angle of phase difference will be 0**. Ans. 



CIRCUITS COISTAINEN^G SBIiF-rNDUCTION ONIiT 

43, In practice, it is, of course, impossible to obtain any 
device or electrical circuit that is entirely devoid of resist- 
ance, so that when we speak of a circuit that contains self- 
induction only, it is understood to be one in which the 
opposition to the current on account of the resistance is 
negligible compared with that of. the induced E. M. F. If 
the circuit has a negligible resistance, it is evident that the 
E. M. F. applied or impressed on the circuit is taken up 
wholly in overcoming the induced E. M. F. ; hence, the 
impressed E. M. F. must be the equal and opposite to that 
induced in the coil. Of course, the only E. M. F. that is 
actually present is that impressed on the circuit, but in 
working alternating-current problems it makes matters 



Digitized by VjOOQ IC 



33 



ALTERNATING CURRENTS 



§16 



I 



clearer to imagine the impressed E. M. F. as opposed by 
counter E. M. F.'s, due to the resistance or inductance, as 
the case may be. It has already been mentioned that when 
an alternating E. M. F. is applied to an 
inductive circuit, the current does not 
respond at once to the changes in the 
E. M. F., but lags behind it, because 
of the opposition that the self-induction 
offers to any change in the current. 

The E. M. F. necessary to overcome 
self-induction is at right angles to the 
current and 90° ahead of the current in 
phase. This may be shown by refer- 
ring to Fig. 26. Let the line oa and 
the corresponding wave shown by the 
full line represent the current flowing 
in the circuit. The magnetism pro- 
^5 S duced by the current will increase and 
decrease in unison with it, and hence 
may be represented by the light-line 
wave in phase with the current. Now 
it has been shown that the induced 
E. M. F. is proportional to the rate at 
which the magnetism changes, and it 
will be seen from the figure that the 
\ magnetism is changing most rapidly at 

the points a and c where the magnetism 
curve cuts the axis, because there is a 
much greater change between two 
points such as e and c than there is 
between d and d\ It follows, then, 
that when the current and magnetism 
are passing through their zero values, 
the induced E. M. F. is at its maximum value; conse- 
quently, it must be at right angles to the current. 



f* 



// 



// 



44, It is now necessary to determine whether this 
induced E. M. F. is 90° ahead of the current or behind it. 



Digitized by VjOO^IC 



§16 ALTERNATING CURRENTS 33 

When the current is rising in the circuit, the induced E. M. F. 
is always preventing its rise; hence, it may be represented 
by the dotted curve, Fig. 26, 90^ behind the current. By 
examining this curve, the student will see that the induced 
E. M. F. it represents opposes any change in the current. 
This induced E. M. F. may be represented by the line oc^ 90** 
behind oa. The impressed E. M. F. necessary to overcome the 
self-induction will be the equal and opposite of this, and will 
be represented by the line ^^, 90** ahead of oa. The E. M. F. 
to overcome the self-induction is also shown by the dot-and- 
dash curve. The student must keep in mind the distinction 
between the E. M. F. of self-induction and the E. M. F. 
necessary to overcome self induction; the former is 90^ behind 
the current in phase, while the latter is 90° ahead of the 
current. 

46, In the foregoing, it has been assumed that the mag- 
netism is exactly in phase with the current. This is not 
always the case where iron is present, as the hysteresis in 
the iron causes the magnetism to lag a little behind the cur- 
rent. It is true exactly for all circuits containing no iron, 
and sufficiently true, for most practical purposes, for iron 
circuits as well. 

46, Value of Induced E. M. F. — The amount of the 
E. M. F. induced in any circuit or device depends on three 
things — the current, the frequency, and the coefficient of 
self-induction. The coefficient of self-induction L has 
already been explained ; for any circuit or device without 
iron, it has a constant value regardless of the current. As 
already shown, 

^-" lo^ 

where L = coefficient of self-induction expressed in henrys; 
^* = number of magnetic lines set up by a current of 

1 ampere; 
T •=. number of turns through which the lines thread. 



Digitized by VjOOQ IC 



84 ALTERNATING CURRENTS §16 

If the device contains iron in its magnetic circuit, as is 
nearly always the case, the value of L will change with the 
current. For example, the inductance of a coil wound on 
an iron core will change with the current, because with large 
currents the iron becomes saturated. If, however, the iron 
is worked well below the saturation limit, L will be fairly 
constant. From the above we have 

^ = 7xTo-- <«) 

where 9> = flux corresponding to current /; 

T = number of turns with which flux 9 is linked. 

The effective value of the induced E. M. F. may be cal- 
culated as follows, when the inductance L is known : If the 
maximum magnetic flux is ^, this total flux is cut four times 
by the coil during each cycle; i. e., the flux increases from 
zero to ^, then decreases to zero, increases to 9 in the nega- 
tive direction, and finally decreases to zero again. Now, by 
definition, the average volts induced in the coil must be equal 
to the average number of lines of force cut per second, divided 
by 10*. If the coil has T turns and the frequency is n cycles 
per second, the average number of lines of force cut by all 
the turns per second will be 4 ^ Tn^ and the average volts 
induced will be 

^av.=l^ (3) 

or, since the effective volts is equal to 1.11 times the aver- 
age volts^ 

effective volts = is = — — —z^ (4) 

47. The formula giving the relation between the induced 
effective volts E^ the frequency «, the number of turns 7", 
and the flux ^ is important, as it is used repeatedly in con- 
nection with alternator, transformer, and induction-motor 
design. It may be expressed as follows: Whenever a viag- 
netic flux is made to vary through a circuit so as to induce a 
sine E. M. F. , the effective value of the E, M, F, so induced 



Digitized by VjOOQ IC 



§16 ALTERNATING CURRENTS * 36 

is equal to J^.J^If times the product of the maximum flux 4^, the 
number of turns T connected in series^ and the frequency n 
divided by 10^. 

48, If the inductance of a circuit is L henrys, we have, 
from formula 8, 

I X 10" " * 

where ^ is the maximum flux through the coil correspond- 
ing to the maximum current I; L the inductance in henrys; 
and T the number of turns. 

If / is the effective current flowing in the circuit, the 

maximum current must be /V^, because / = — — I = .707 I. 

Hence, — -= = L (6) 

/ 1/2 X 10- ^ ' 

or ^T^ I^LW (6) 

But from formula 4, the effective E. M. F., 

.707^ r — 

^ _ 4.44 ^Tn _ ^ .636 _ , V^ ^Tn 

^ " lO*"^ " W " '^ ^ T" ^ 10" 



NoTB.-rIn the above equation, the coefficient 4.44 = 4 X 1«11 = 4 

.707 ^ ^ 
-^ .686 * '^ 2 • 

IT 

Substituting for 9 T the value given by formula 6, 

1 

£=4X^X«//2Z, 

E^%i:nLI (7) 

This may be expressed as follows : Whenever an alterna- 
ting current of I amperes {effective^ is sent through a circuit 



Digitized by VjOOQIC 



d6 



ALTERNATING CURRENTS 



16 



of inductance L henry s^ the value of the induced E. M. F. , in 
effective volts, is equal to 2 it times the product of the fre- 
quency «, the inductance Z, and the effective current L 

In some works on alternating currents the quantity ^-k n 
appearing in formula 7 is denoted by the Greek let- 
ter GO (omega), so that 2w«Z is written gjZ, and formula 7 
reads £ = oo L L 

49. Reactance, — It will be seen from the foregoing that 
in order to obtain the volts necessary to overcome self-induc- 
tion, the current is to be multiplied by the quantity 2 r « Z. 
In order to obtain the E. M. F. necessary to overcome 
resistance, the current / is multiplied by the resistance R, 
It follows, then, that the quantity ^-kuL is of the same 
nature as a resistance and is used in the same way as a 
resistance. The quantity 2 ^ « Z is called the reactance 
of the circuit, and, like resistance, is measured in ohms. 

The reactance of any inductive circuit is equal to 2 ;r times 
the frequency times the inductance Z expressed in henrys. 

Later on it will be seen that a more general definition of 
reactance as applied to any circuit is that it is the quantity 
that, multiplied by the current, gives the component of the 
impressed E. M. F. that is at right angles to the current. 

60. Take the example shown in Fig. 27. The alterna- 
tor A is connected to a coil or circuit a b^ as shown, the 




T 

I 
i 

^ 






FlO. 27 



inductance of which is .5 henry. The resistance of the cir- 
cuit will be considered as negligible. The frequency of the 
E. M. F. furnished by the alternator is 60. Required, the 



Digitized by VjOOQ IC 



§16 ALTERNATING CURRENTS 87 

E. M. F. necessary to force a current of 3 amperes through 
the circuit. The induced counter E. M. F. will be 

£=2?r«Z:/=2rx60X.5x3 = 565 volts 

Hence, in order to set up the current of 3 amperes, the alter- 
nator must furnish an E. M. F. of 665 volts. This example 
shows one difference between the behavior of alternating cur- 
rents and direct currents. The E. M. F. required to set up 
a continuous current of 3 amperes through a circuit of very 
small resistance, such as this, would be exceedingly small, 
whereas it requires an alternating E. M. F. of 565 volts to 
set up the same current. In other words, if a continuous 
pressure of 565 volts were applied to the coil, the resulting 
current would be enormous. Instances of this effect are 
met with very commonly in connection with transformers. 
The primary coil of a transformer has a high self-inductance 
when the secondary is not loaded, so that, when the primary 
is connected to the alternating-current mains, only a very 
small current flows. If the same transformer were con- 
nected to continuous-current mains, it would be at once 
burned out, on account of the large current that would flow. 

61, In the above example, the circuit was supposed to 
have negligible resistance, so that the whole of the applied 
E. M. F., 565 volts, was used to overcome the self-induc- 
tion. This being the case, it follows that the E. M. F. 
applied by the alternator must be 90° ahead of the current. 
The state of affairs existing in the circuit may therefore 
be represented as in Fig. 28. Lay off to scale the value of 
the current o a (so many amperes per inch) in the direction 
of the line o x. The induced E. M. F. = 2 jt « Z / = 565 volts 
must then be represented by the line o b' (so many volts per 
inch) 90"^ behind oa, and the equal and opposite of this 
o by 90° ahead of o a, will represent the E. M. F. that the 
alternator must supply to overcome the E. M. F. of self- 
induction. Of course, in actual practice it is impossible to 
obtain a circuit that has no resistance, as assumed in the 
above example ; but if the reactance liznLx^ very large as 



Digitized by VjOOQ IC 



ALTERNATING CURRENTS 



§16 



compared with the resistance, a condition quite frequently 
met with, the effect obtained in the circuit will be quite 



•I 






9^mp. 



Pig. 28 



closely represented by Fig. 28, and the current and E. M. P. 
will be nearly at right angles to each other. 



CIRCUITS COKTATNINO RESISTANCE AND SBIiF- 
INDUCTION 
62, Components of Applied E. M. F. — It has been 
shown that the effect of self-induction is to choke back the 
current. It also makes the circuit act as if it possessed 
inertia, as the current does not respond at once to the 
changes in the applied E. M. F., and thus lags behind. The 
resistance of the coil also tends to prevent the current from 
flowing, but it does not tend to displace the current and 
E. M. F. in their phase relations. In considering the flow 
of current through circuits containing resistance and self- 
induction, it is convenient to think of the resistance and 
self-induction as setting up counter E. M. F.'s, which are 
opposed to the E. M. F. supplied by the alternator. The 
E. M. F. supplied from the alternator or other source must 
then, in the case of alternating-current circuits, over- 
come not only the resistance, but also the self-induction. 
In the case of continuous-current circuits, the resistance 
only must be taken into account. In every case, then, 
where an impressed E. M. F. encounters both resistance 



Digitized by VjOOQ IC 



§16 



ALTERNATING CURRENTS 



and self-induction in a circuit, it may be looked on as split 
up into two components, one of which is necessary to over- 
come the resistance and the other the self-induction. 

Take, for example, the case shown in Fig. 29. The 
alternator A is supplying an E. M. F. E. that is setting 

Current^ I 




jlnduetance'L 



Pio. 20 

up a current / through the circuit a b. This circuit pos- 
sesses both resistance R (ohms) and inductance L (henrys). 
In such a circuit it might be desired to find the impressed 
E. M. F. necessary to set up a given current ; or, given the 
impressed E. M. F., to find the current. Suppose it is 
required to find the impressed E. M. F. E necessary to set 
up a given current /. The E. M. F. required must be the 




B.I 



&^Tt^9^^ 



Pio. 80 



resultant sufn of the E. M. F. necessary to overcome the resist- 
ance and that required to overcome the self-induction. The 
former must be equal to /?/, and must also be in phase 
with the current, while the latter is equal to the product of 
the current and reactance I-k n LI and must be 90° ahead 
of the current. In Fig. 30, ^ ^ is therefore laid off to repre- 
sent RI in the same direction as the current, and ob, 
90° ahead of the current, is laid off equal to 2 tc n LI. The 
impressed E. M. F. E must be the resultant oi oa and ob^ 



Digitized by VjOOQ IC 



40 



ALTERNATING CURRENTS 



§16 



or the diagonal oc^ the electromotive forces being com- 
bined according to the parallelogram of forces, as previ- 
ously explained. The diagonal o c therefore represents the 
E. M. F. set up by the alternator to the same scale that o a 
and^^ represent the component E. M. F.*s. The diagram 
also shows that the current lags behind the impressed 
E. M. F. by the angle 0, and it is also seen that if the circuit 
had no inductance, the line o b would become zero, and the 
current would be in phase with the E. M. F. 

53. Instead of using a complete parallelogram, as in 
Fig. 30, triangles are commonly used to show the relations 




B/ £,MJF. to overeotne ResUtanci. 
PlO. 81 

between the different E. M. F. 's. The right triangle, Fig. 31, 
shows the same relations as the parallelogram, Fig. 30; 
oa = RI represents the E. M. F. necessary to overcome 
resistance, ac that to overcome self-induction, and oc is the 
resultant. It should be remembered that the line ac is 
transferred from o b, and consequently represents an E. M. F. 
90** ahead of o a. Since the angle oac is a right angle, it 
follows that 



or 






(8) 
(9) 

(10) 



or E = /i^R'+{2i:nLy 

That is, the impressed E. M. F. E necessary to maintain a 
current /in a circuit of resistance R and inductance L is equal 
to the product of the current /into the square root of the sum 
of the squares of the resistance R and the reactance 2 i^nL. 



Digitized by VjOOQ IC 



§16 



ALTERNATING CURRENTS 



41 




Xe§i8tanee mB 

Pio. m 



The quantity V^ + (2w«Z)* is called the impedance of 
the circuit. The impedance of a circuit is equal to the square 
root of the sum of the squares of the resistance and reactance. 

64, The relation between resistance, reactance, and im- 
pedance is shown by the right-angled triangle, Fig. 32. The 
following definitions may 
also be given of impedance, 
reactance, and resistance: 

Impedance is that quan- 
tity which multiplied by 
the current gives the im- 
pressed E. M. F. 

Reactance is that quantity which multiplied by the cur- 
rent gives the component of the impressed E. M. F. that is 
at right angles to the current. 

Resistance is that quantity which multiplied by the cur- 
rent gives the component of the impressed E. M. F. that is 
in phase with the current. 

Impedances, like resistances and reactances, are expressed 
in ohms. 

55. In dealing with continuous-current systems, the 

relation between the current, resistance, and E. M. F. is 

E M F 

fully given by Ohm's law, i. e., current = — ^— -^ . Ohm's 

^ ^ '' resistance 

law cannot, however, be applied in this form to alternating- 
current circuits, but from formula 10 becomes 



/= 



E 



or 



j^R'-^-ClitnLf 

E. M. F. 



(11) 



current = r 



impedance 



and the current no longer depends simply on the E. M. F. 
and resistance, but depends also on the inductance L and 
the frequency n. If ;/ becomes zero, i. e., if the alterna- 
tions become slower and slower until the current finally 
becomes continuous, the term (2 n n Ly drops out and 



Digitized by VjOOQ IC 



ALTERNATING CURRENTS 



§16 
Also, if the 



E E 

formula 11 reduces to / = — ;== or / = ^. 

circuit is non-inductive, i. e., if Z = 0, the formula reduces 
to Ohm's law, and the current may be obtained by simply 
dividing the E. M. F. by the resistance, as is the case with 
continuous currents. 



AKGIiB OF ULG 

66. From the triangle, Fig. 32, it will be seen that the 
current that is in the direction oi o a lags behind the 
impressed E. M. F., in the direction of ob^ by the angle 0. 

The tangent of this angle is equal to — ^ — J hence, if the 

resistance, inductance, and frequency are known, the angle of 

^"K n L 
lag can be calculated. Prom the relation tan ^ = — = — , 

it is seen that the larger 2 w ;i Z is, compared with R^ the 
larger will be the angle of lag, and if the reactance 2nn L 
is small in comparison with R^ the angle <t> will be small, 

or. the current will be nearly in 
phase with the impressed E.M.F. ; 
hence, in a circuit containing 
resistance and self-induction, the 
current lags behind the impressed 
E. M. F. , the amount of the lag de- 
pending on the relative magnitude 
of the resistance and reactance. 

Example. — An alternator is con- 
nected to a circuit having a resistance of 
20 ohms and an inductance of .1 henry. 
The frequency is 60 cycles per second. 
What must be the E. M. F. furnished 
by the alternator in order to set up a 
current of 10 amperes in the circuit ? 

Solution. — The required E. M. F. 
must be the resultant of the E. M. P. 
necessary to overcome resistance, i.e., 
y? / = 20 X 10 = 200 volts, and that 
necessary to overcome the self-induc- 
tion. The latter is equal to the react- 
ance multiplied by the current, or 2ir n LI = ^X 3.14 X 60 X .1 




Digitized by VjOOQ IC 



§16 ALTERNATING CURRENTS 43 

X 10 = 876.8 volts. The impressed E. M. F. E is the resultant, or 
E— f^200* -f 876. 8« = 426.5, nearly. These E. M. F.'s are related as 
shown in Fig. 88, and the resultant E may either be found by calcula- 
tion, as shown, or it may be scaled from the figure. The required 
E. M. F. that must be supplied by the alternator is therefore 426.5 volts. 
The current in the circuit will lag behind the E. M. F. by an angle ^, 

2 TT f* / 

the tangent of which is — ^ — = l-^^- ^Y looking up the angle in a 

table of tangents, it is found to be a little over 62". This means, then, 
that in this particular circuit the current does not come to its maximum 
value until a little over \ period behind the E. M. F. Since the current 
passes through 60 cycles per second, it follows that the current in this 
case rises to its maximum value about jj© second after the E. M. F. 

The impedance of the circuit is 4/20*-+- 37.68* = 42.65 ohms, and this 
multiplied by the current gives 426.5 volts as the impressed E. M. F. 

Ans. 

67. The foregoing problem might also come up in an-" 
other form: The impressed E. M. F. being given, to deter- 
mine the current. From formula 11, 

E 



/= 



so that, if E is given, / can easily be determined, R and L 
being known quantities. 

If there were no inductance in the circuit, the E. M. F. 
required would be in accordance with Ohm's law, or 20 X 10 
= 200 volts, which is less than half the voltage required 
when the inductance is present. 

If there were no resistance present, the tangent of the 

angle of lag would be — - — = 00, or the angle of lag 

would be 90°, all the impressed E. M. F. being used in over- 
coming the inductance. 

CIRCUITS CONTAINING CAPACITY ONI.T 

68, If an electrical condenser be connected across the 
terminals of an alternator, we have an example of a circuit 
of large capacity, and as the resistance of the connecting 
wires is very small, the resistance of the circuit outside of 
the condenser itself is practically zero. We then have a 
circuit that may be*considered as containing capacity only. 

44—26 



Digitized by VjOOQ IC 




44 ALTERNATING CURRENTS §16 

69. Condenser Cliargres. — If a battery be connected to 
the terminals of a condenser, as shown in Fig. 34, a current 

will flow into it and the plates will 
become charged. The flow of cur- 
rent will be a maximum the instant 
the E. M. F. is applied, but will 
rapidly fall off, so that in a small 
fraction of a second the current will 
practically have ceased flowing and 
the condenser will be charged. 
This will be the state of affairs so 
long as the condenser remains con- 
nected to the battery ; except for the 
instant when the battery is first connected, no current will 
flow, and the circuit will act simply as if it were broken. 
The condenser acts as if it had acquired a counter E. M. F., 
tending to keep out the current, and this counter E. M. F. 
becomes greater until, when the condenser is charged, it is 
equal and opposite to that of the battery. If the battery be 
disconnected and the terminals of the condenser connected 
together, the charge will flow out and will result in a cur- 
rent of short duration. This current will be a maximum 
when the terminals are first connected, but it soon falls to 
zero. The unit of capacity is the farad, which has already 
been defined. For convenience, the microfarad is used as 
the practical unit, but in working out problems, capacities 
must always be expressed in farads before substituting in 
formulas, because the farad is chosen with respect to the 
volt and ampere, and hence must be used in formulas 
together with these units. For example, a capacity of 
10 microfarads as given in a problem would be substituted 
in formulas as .00001 farad. 

60, In connection with condensers and capacities, it is 
often necessary to make use of the unit denoting quantity of 
charge. The unit of quantity is the co?4lomd, which has already 
been defined. It will be convenient to repeat here some of 
the definitions given, in order to assist in a thorough compre- 
hension of that which follows. The coulomb represents that 



Digitized by VjOOQ IC 



§16 



ALTERNATING CURRENTS 



45 



quantity of current that passes through a circuit when the 
average rate of flow is 1 ampere for 1 second. II Q= quantity 
of electricity, or charge, in coulombs that a condenser takes 
up in / seconds, the average current during the time / must 
have been 



^=7 



(12) 



61. If a condenser C, Pig. 35, be connected to an alter- 
nator, a current will flow into and out of it, because the 
E. M. F. at its terminals is constantly changing. If an 





Pio. » 

ammeter M is connected in the circuit, it will give a reading 
just as if the alternator were sending a current through an 
ordinary circuit, whereas there is really no electrical con- 
nection between the terminals of the condenser C, and if 
it were connected to a continuous-current dynamo, the 
ammeter Af would give no reading whatever. What really 
occurs is a surging of current into and out of the condenser. 




63. Condenser B. M. F. — Let the full-line curve, Fig. 36, 
represent the current that flows into and out of a condenser 
when a sine E. M. F. is impressed on its terminals. This 
current will, of course, have a frequency equal to that of the 



Digitized by VjOOQ IC 



46 ALTERNATING CURRENTS §16 

alternator to which the condenser is attached. It will require 
a certain impressed E. M. F. to cause this current to flow, just 
as it required an E. M. F. to overcome the inductance of a 
circuit. The problem now is to determine the value of this 
E. M. F. and its phase relation with regard to the current. 
The E. M. F. required to set up the current will be the 
equal and opposite of the E. M. F. of the condenser, just as 
the E. M. F. to overcome self-induction was the equal and 
opposite of the E. M. F. of self-induction. 

63. It has already been shown that when the flow of 
current into the condenser is a maximum the counter 
E. M. F. of the condenser is zero, and when the flow finally 
becomes zero the counter E. M. F. is a maximum. It fol- 
lows, therefore, that the wave representing the E. M. F. of 
the condenser is at right angles to the current. The fact as 
to whether it is 90° ahead of or behind the current may be 
decided as follows: When the current is flowing into the 
condenser, the counter E. M. F. is continually increasing in 
such a direction as to keep it out. The curve representing 
the E. M. F. of the condenser must cross the axis at the 
point d. Fig. 36, because, as shown above, this curve is at 
right angles to the current curve. During the interval of 
time from b to ^, the current is decreasing, so that during 
this interval the counter E. M. F. of the condenser must be 
increasing in the opposite direction, and is therefore repre- 
sented by the dotted curve, which is 90° ahead of the cur- 
rent. The impressed E. M. F. necessary to overcome that 
of the condenser must be the equal and opposite of this, 
or 90° behind the current. This latter E. M. F. is shown 
by the dot-and-dash curve. The lines in the diagram at 
the left show these phase relations, the line o a representing 
the current, oc the E. M. F. of the condenser, and ob the 
impressed E. M. F. to overcome that of the condenser. 

Hence, in a circuit containing capacity only, the current 
is 90° ahead of the impressed E. M. F., or the E. M. F. 
necessary to set up a current in such a circuit is 90° behind 
the current in phase. 



Digitized by VjOOQ IC 




Pio. 87 



§16 ALTERNATING CURRENTS 4» 

In Fig. 37, the current is represented by the line Oa and 
the impressed E. M. F. necessary to maintain the current 
by Ob, 90° behind O a. The 
counter E. M. F. of the condenser 
itself would be represented by Oc^ 
equal and opposite to O b^ and 

hence 90° ahead of the current. „■ 

^ \ ^** 

64. The student will note from 

the foregoing that the effect of ca- 
pacity in a circuit is exactly the 
opposite of self-induction. Capac- 
ity tends to make the current lead 
the E. M. F., while self-induction causes it to lag. When 
both self-induction and capacity are present in a circuit, one 
tends to neutralize the other. 

66. The E. M. F. in volts (effective) necessary to over- 
come the capacity or condenser E. M. F. in a circuit may be 
calculated as follows: 

If the capacity of the condenser be C farads and a maxi- 
mum E. M. F. E be applied to its terminals, it will take up 
a maximum charge g = C E coulombs. The E. M. P. 
passes through n cycles per second, i. e., the condenser is 
charged up to a maximum in one direction, then discharged, 
and the process repeated in the opposite direction, n times 
per second. The average rate of charge and discharge is, 
therefore, 4 «, i. e. , 4 « times per second. 

The maximum rate = average rate X ^; hence, themaxi- 

At 

mum rate of charge and discharge is — ^r — = 2 w «. The 

maximum charge is C E, and if the maximum rate of charge 
is 2 7r«, the maximum current must be 

I = 2^«CE (13) 

This formula gives the relation between the maximum 
E. M. F. and maximum current. The effective E. M. F. is 



Digitized by VjOOQIC 



48 ALTERNATING CURRENTS §16 

equal to the maximum divided by |/2. Dividing each side 
of the equation by |/2, we have 

so it is seen that this equation also gives the relation between 
the effective E. M. F. and current. 

66. Capacity Reactance. — The quantity yr is called 

the capacity reactance, as it is analogous to the react- 
ance UtthL in circuits where self-induction is present. It 
has also been called the condensance by some writers. 

Example 1. — Required, the E. M. F. necessary to set up an alter- 
nating current of 2 amperes through a condenser having a capacity 
of 5 microfarads, the frequency being 60 cycles per second. 

Solution. — By formula 15, we have 

7=2 amperes ; C = 6 microfarads = .000005 farad ; « = 60; 

^= 2 X 8.14 X 60 X .000005 = ^'^^ ^°'^ ^"* 

Example 2. — B By Fig. S8, represents a high-tension power-transmis- 
sion line connected to an alternator A, The pressure maintained be- 
tween the lines is 10,000 volts, and the frequency is 60 cycles per second. 



Current 1.5 ampef. 




Pig. 88 



There is no connection between the wires, and they are supposed to be 
so insulated that practically no leakage takes place between them. On 
running the alternator, it is found that the ammeter M gives a read- 
ing of 1.5 amperes. What must be the electrostatic capacity of the line ? 
Solution.— From formula 16, we have, by derivation, 

^^ UrUTE ^ 2 X 3.14 X 60 X 10,000^^*^ 

1.5X1,000,000 ooQ • * ^ A 

— - = .898 microfarad Ana. 



'2X8.14X60X10,000 



Digitized by VjOOQIC 



ALTERNATING CURRENTS 

(PART 2) 



BESISTANCE, SELF-rNDUCTION, AND 
CAPACITY 



CIRCUITS CONTAINING RESISTANCE AND 
CAPACITY 

1. In case a circuit contains resistance and capacity, the 
current will lead the E. M. F. The amount of lead will 
depend on the relative values of the resistance and capacity 
reactance. If the resistance is very large compared with 
the capacity reactance, the angle of lead will be small, 
because that component of the E. M. T. at right angles to 
the current will be small. On the other hand, if the capac- 
ity reactance is very large compared with the resistance, 
the current may lead the E. M. F. by nearly 90°. 

2. The resultant E. M. F. may in such circuits be looked 
on as being composed of two components, one at right 
angles to the current, used in overcoming the capacity 

reactance and equal to y^ and the other, necessary to 

overcome the resistance, equal to R I and in phase with the 
current. These E. M. F.'s may be represented by the 

§17 

For notice of copyright^ see page immediately following the title page. 



Digitized by VjOOQ IC 



ALTERNATING CURRENTS 



§17 



diagram, Fig. 1. ox represents the current and oa -= RT 
the E. M. F. to overcome resistance. The E. M. F. to 



€&vrTtnt . 




Pio. 1 

overcome the capacity reactance is represented hy ob 
90° behind (7 tf, and the resultant impressed E. M. F. -£by the 
diagonal o c. The current leads the E. M. F. by the angle 0, 

/ 

the tangent of which is equal to ^^ = — p-7 — = p^ 



oa 



^"^ = 2;r^z^ 



-irr 
(1) 



3* -From formula 1 it will be noticed that if R becomes 
zero, tan ^ = infinity, or the angle of lead is 90°. If the 
capacity C becomes infinitely large, tan ^ = and the cur- 
rent is in phase with the E. M. F. This latter is the condi- 
tion of affairs in an ordinary closed circuit, because in such 
a case the current keeps on flowing so long as the E. M. F. 
is applied, and the circuit never becomes charged. In other 
words, an ordinary closed circuit in which there is no con- 
denser at all acts as if it had an infinitely large capacity, 
while ah open circuit acts as a condenser of infinitely small 
capacity. 

4* From the triangle oac^ Fig. 1, we have the relation 



£" = R'/' + 



= \/^' 



/' + 



4 7rV/'6^* 



4 7r««»C« 



(3) 



Digitized by LjOOQIC 



17 



ALTERNATING CURRENTS 



8 



^=V'''+(^7 



(3) 



Therefore, in circuits containing resistance and capacity^ the 
impedance of the circuit is equal to the square root of the sum 
of the squares of tlie resistance R and the capacity reactance 
1 

The law governing the flow of current in such a circuit 
then becomes 

E 



/ = 



V 



^+ 



4»'«'C' 



(4) 



If the circuit contains a resistance R and no condenser, 

C becomes infinite in value, and formula 4 reduces to 

E 
/ = -^. That is, the current follows Ohm's law. 

If the circuit be broken, it means that the resistance 
becomes infinitely large, and the capacity C being very 
small, the impedance becomes infinitely large; consequently, 
the current becomes zero. 




T 

■9 



Pio. d 




Example. — A non-inductive resistance R^ Fig. 2, of 200 ohms is 
connected in series with a condenser across the terminals of an alter- 
nator, as shown, the frequency being 60. The condenser has a capacity 



Digitized by VjOOQ IC 



ALTERNATING CURRENTS 



§17 



of 16 microfarads, and the current flowing in the circuit is 5 amperes. 
Required: 

1. The reading that would be given by the voltmeter Ki connected 
to the terminals of the resistance. 

2. The reading of the voltmeter F, connected to the terminals of 
the condenser. 

8. The angle by which the current will lead the E. M. F. 
4. The E. M. F. that must be furnished by the alternator, i. e., the 
reading of the voltmeter V connected across the mains. 

Solution. — We know that the three required E. M. F.*s must be 
related to each other as shown in Fig. 1. 

1. The reading given by the voltmeter Vi must evidently be the 
E. M. F. necessary to overcome the resistance R, and hence is equal to 
RI= 200 X 5 = 1,000 volts. Ans. 

2. The reading of V% represents the E. M. F. necessary to overcome 
the capacity reactance, and hence is equal to 



2fl'«C " 2 X 8.14 X 60 X .000015 



= 884 volts Ans. 



8. The angle by which the current leads the E. M. F. is given by 
formula 1 ; tan ^ = ^ j^-^ = .884. From a table of tangents ^ 

ATT H L K 
is found to be 41* 29', nearly ; i. e., the current is ahead of the E. M. F. 
by a little more than one-ninth of a complete cycle. Ans. 



BI ^1000 VolU 




PlO. 8 

4 The resultant E. M. F. E, or the voltage that must be furnished 
by the alternator to set up the current of 5 amperes, is obtained from 



Digitized by VjOOQIC 



8 17 



ALTERNATING CURRENTS 



formula a : iS" = J^ R} /« -h y ^,^, ^, = i^l,000« -h 884« = 1.886 volts. 

This is the pressure that would be given by the voltmeter K, and it is 
the resultant sum of the E. M. F. to overcome resistance (1,000 volts) 
and that to overcome reactance (884 volts), 

6* In an alternating-current circuit as considered in the 
above example, it is thus seen that it is quite possible for 
the reading of the voltmeter V to be considerably less than 
the arithmetical sum of V^ and F,. 

The relation between the quantities in this problem is 
shown by the E. M. P. triangle, Fig. 3, the sides of the tri- 
angle being laid off to scale to represent the different 
E. M. F.'s. 




6, Resistance and Capacity In Parallel. — ^An example 

of resistance and capacity connected in parallel is shown in 

Fig. 4. The alternator 

mamtains a constant 

E. M. F. E across the 

terminals of both R 

and C, The current / 

that will flow in the 

main circuit under such 

conditions is determined as follows: Let the line o a, Fig. 6, 

represent the impressed E. M. F. E. i? is a non-inductive 

resistance ; hence, 
the current in 
branch 1 will be /, 

= -j^ and will be in 



PIO. 4 




..* 



Fig. 5 



phase with £; hence, 
it may be represented 
by a line such as o b. 
The current in branch )& is /, = 2 ^ « C -£, and is 90° ahead 
of the E. M. F. in phase. The current in the line must be 
the resultant of /, and /,, and is, therefore, represented by 
/, which l eads the E. M. F. by the angle and is equal to 



Digitized by VjOOQIC 



ALTERNATING CURRENTS 



1 17 



CIBCUrrS containing SEIiF-INDUCnON AND 
CAPACITY 

7. When a circuit contains self-induction and capacity, , 
the two tend to neutralize each other, because it has already 

been shown that the effect of 
capacity is precisely the opposite 
to that of self-induction. Take 
the case shown in Fig. 6 where an 
inductance L and capacity C are 
connected in series across a circuit. 
Let / be the current that flows in 
E. M. F. E^ across the inductance 



r 




Pio. e 
the circuit. Then the 
will be 2 TT « Z /, and will be 90** ahead of the current, or, in 
other words, the current 
will lag 90° behind the ^ 

E. M. F. £,. T h e -^^i - 2nrnLl 

E. M. F. E^ across 
the capacity C must be 

^•=2ir^' ^"^ ^"^^ 

current / must be 90** 
ahead of the E. M. F. E^. 
This state of affairs may 
be represented as shown 
in Fig. 7; ^ « represents 
the current /, and o b the 
E. M. F. £. = 2 r « Z /, 

90** ahead of the current \ o c\s equal to E^ = ^, and 



«• = 



2'KnC 



E 



Piaf 



is 90° behind the current, or the current / is 90° ahead of 
the E. M. F. across the condenser. The line E. M. F. E is 
the difference between o c and ob\\\. is, therefore, represented 
by d^ found by subtracting a length dc = ob from oc. 
The line E. M. F. may be either ahead of or behind the 
current in phase, depending on the relative effects of the 
self-induction and capacity. It is easily seen that uVider 
the conditions shown in Fig. 7, the pressures E^ and -£, may 



Digitized by VjOOQ IC 



§17 



ALTERNATING CURRENTS 



be very much higher than the line pressure. It is, of course, 
impossible to get a 
circuit absolutely 
devoid of resistance, 
so that in practice 
b and o c would not 
be exactly at right 
angles to o a. The 
conditions that would 
actually arise in 
practice would be 
more nearly repre- 
sented by Fig. 8, 
where o b and o c are 
not exactly at right 
angles to o a, and 
where the resultant 
E. M.F.od is found 
by completing the 
parallelogram obdc. 



Pio. 8 





PIO. 



8. SelT-Inductlon and Capacity In Parallel. — Fig. 9 
shows an inductance and capacity in parallel. In this case 

Z, C, and E are known, 
/ I " and it is desired to find 

the value of /. The 
current in L would be 

^"~ / = =r, since there 

is supposed to be no resistance present. The current in 
C would be /^ = 27:nC£. /, is 90° behind the E. M. F., 
and /„ 90° ahead of the E. M. F., so that the diagram 
for the circuit will be as shown in Fig. 10. oa repre- 
sents the E. M. F. E; ob the current /,, and oc the cur- 
rent /,. The resultant current is / and shows that the 
current in the main line may be much less than the 
currents in the separate branches. If some resistance 
were present in each branch, /, and /, would not be 



Digitized by VjOOQ IC 



8 ALTERNATING CURRENTS § IT 

exactly at right angles to £, and the diagram would be 



d4 ^ 



Pio. 10 Pio. 11 

as shown in Fig. 11, od =: I being the current supplied 
by the alternator. 



CIRCmTS CONTAINING BBSISTANCB, SITLF- 
INDUCnON, AND CAPACITY 

9* Resistance, Self-Indnetlon, and Capacity In Series. 

It quite frequently happens that a circuit may contain all 
three of these. The effect of all these three quantities — 
resistance, self-induction, and capacity — being present in a 
circuit is easily understood if it is remembered that self- 
induction and capacity always tend to neutralize each other. 
Suppose an alternator A^ Fig. 12, is supplying current to 
the circuit a b containing resistance R^ inductance Z, and 

capacity C, and suppose that the capacity reactance j^ 

is less than the reactance ^nn L due to the self-induction. 
The resultant E. M. F. E necessary to maintain the current 
in the circuit may then be determined as shown in Fig. 13. 
o a represents, as in previous problems, the E. M. F. neces- 
sary to overcome the resistance = R I and in phase with the 



Digitized by VjOOQ IC 



§17 



ALTERNATING CURRENTS 



9 



current along the line ox\ oc^ the E. M. F. to overcome 
self-induction, 90** ahead of oa^ equal to %'Kn LI\ and ob^ 




RetUtanee Jt 



Inductance L 



OagacUp C 



Pio. 13 




Cttrreni 



Pio. 18 



the E. M P. necessary to overcome capacity, 90** behind oa^ 
equal to rrr. The total component at right angles to 



Digitized by VjOOQIC 



10 ALTERNATING CURRENTS §17 

the current will be ^izn L I — ; tt, and may be laid 

off by measuring back from c^o' V =■ ob, oe is then the 
resultant component at right angles to the current, and 
the impressed E. M. F. furnished by the alternator is the 
resultant of ^^ and o a^ i. e., the diagonal o d. This result- 
ant E. M. F. is in this case ahead of the current by the 
angle 0, or the current is lagging behind the E. M. F. 
From the triangle o da^ Fig. 13, we have the relation 

^• = /'[^'+(2.«Z-^y] (6) 

and n = I^R^^{^.nL-^^ (6) 

The expression under the square root sign is, therefore, 
the impedance of a circuit possessing resistance, self-induc- 
tion, and capacity, because it is that quantity which multi- 
plied by the current gives the E. M. F. 

The impedance of a circuit containing resistance ^ self -indue- 
tion^ and capacity is equal to the square root of the sum of 
the squares of the resistance and the difference between the 
reactance due to self-induction and the reactance due to 
capacity. 

From formula 6, we have 

giving the relation between current and E. M. F. for a cir- 
cuit containing all three of the above quantities. 

10. The tangent of the angle between the E. M. F. and 
current is given by the expression 

2 r ;/ Z. ~ ^ 



tan = ^ (8) 



Digitized by VjOOQIC 



17 



ALTERNATING CURRENTS 



11 



So long as the expression "Zt: nL — -^ is positive, i. e., 

Aft tl C 

when the self-induction has a greater effect than the capac- 
ity, the component oe^ Fig. 13, will be above oa and the 
current will lag behind the E. M. F. If the expression is 
negative, i. e., when the capacity has a greater effect than 
the self-induction, the component o e will be negative, or 
below the line o x, and the current will lead the E. M. F. In 
such a circuit, therefore, the angle between the E. M. P. 
and current may vary 90° either way. Fig. 14 shows a case 




PIO. 14 



where the capajcity has a greater influence than the self- 
induction, i. e., ^^ is less than ob. Measuring o' V = ob 

I 



from c gives ^ ^ as the difference = 2 ?r « Z / 



-, and 



2nnC 

the E. M. F. £ = ^ d. The current in this case is ahead 
of the E. M. F. by the angle ^. 

1 



11. When 2w;iZ = 



%nnC' 



the expression 2 tc n L 



— 7T becomes zero, and = 0. When this is the case, 

2 TT w c * 

the current is in phase with the E. M. F. and follows Ohm's 
law, the capacity and self-induction neutralize each other, 



44—27 



Digitized by VjOOQ IC 



12 ALTERNATING CURRENTS §17 

and, though both are present in the circuit, the effect is the 
same as if neither were there. 

In practice, it is almost impossible to obtain complete 
neutralization of self-induction by capacity ; but if the con- 
ditions are favorable, it may be approached quite closely. 

12. If the self-induction were neutralized by the capacity, 
the current flowing under a given impressed E. M. F. would 
be a maximum and would be determined by Ohm's law. 

This condition is shown in Fig. 15, where (? ^r and (? ^ are 
equal and opposite, and R I ^ E because the current and 
E. M. P. are in phase. 



s,« 



u 



M^BI 



Pio. 15 



It should be noted that, with given values of self-induction 
and capacity, this condition can exist only for one particular 
value of the frequency, and for any other value the two would 
not neutralize each other. If 

""""^^^ 

we have n = ^J j^ = ^isj h, <^> 

which gives the value of the frequency corresponding to the 



Digitized by VjOOQIC 



§17 ALTERNATING CURRENTS 13 

values of C and L. This production of a maximum current 
in a circuit for a certain criticar value of the frequency is 
known as electrical resonance. 

13« Resonance sometimes produces peculiar effects in a 
circuit. If the resistance of the circuit is very low, a neu- 
tralization of the self-induction would allow a large current 
to flow. The E. M. F. across the terminals of the con- 
denser is ^, and if / be very large, this pressure may 

rise to a value very much greater than the impressed E. M. F. 
In the case of an underground cable, the pressure between 
the wire and the sheath might rise sufficiently to break down 
the insulation, while at the same time the E. M. F. supplied 
by the generator might not be at all high. Usually, however, 
the frequencies employed in practice are not high enough to 
make the effects of resonance very common. 

14, The following example will show the application of 
the above formulas to a circuit containing resistance, self- 
induction, and capacity : 

Example. — An alternator A, Fig. 16, is connected to a circuit 
having resistance R = 100 ohms, self-induction Z = .25 henry, and 



I h— -»j — -4 — ^ — t— ^^— 




PIO. 16 

capacity C = 20 microfarads. The current flowing is 5 amperes, and 
the frequency 60 cycles per second. » 

1. Find the E. M. F. or drop Ei across the resistance, drop £% 
across the inductance, drop Et across the condenser. 

2. Determine whether the current lags behind the impressed 
E. M. F., or is ahead of it, and by what amount. 

8. Find the value of the impressed E. M. F. ^necessary to maintain 
the current of 5 amperes. 



Digitized by VjOOQ IC 



14 ALTERNATING CURRENTS §17 

Solution. — 1. If the inductance Z = .26 henry, the reactance 
2vnL = 2x8.14X«0x .25 = ©4.2 ohms. 

The capacity C = 20 microfarads, hence the capacity 
1 1,000.000 .^„ . 

reactance = ^^^^ = 2x3.14x60x20 = ^^'^ ^^™^- 

The E. M. F. necessary to overcome resistance := R I 
= 5 X 100 = 500 volts = El. 

The E. M. F. necessary to overcome inductance = IX%'iTnL 
= 5 X 94.2 = 471 volts = ^,. 

The E. M. F. necessary to overcome capacity = ^ ^ 

= 132.7 X 5 = 663.5 volts = Et. Ans. 

2. Since the E. M. F. necessary to overcome capacity is 
663.5 volts and is 90" behind the current, while that to over- 






^< 



? 



I 



kK'- 



t 

I 



BcaU l'^2SO roll#. 



Pio. 17 



come self-inductance is only 471 volts and 90" ahead of the 
current, the resultant component at right angles to the 

current is 2iTnLI - s tt = 471 - 663.5 = — 192.5, and is 

90" behind the current (the negative sign indicating that this 
component is below the current line). It follows, then, that 
the impressed E. M. F. E lags behind the current, or the cur- 
rent is in advance of the E. M. F. The relation of the different 
E. M. F.*s will be readily seen by referring to Fig. 17. The angle by 
which the current leads is easily found from the figure, because 

192 5 
^^ ^ — ri\(\ — -^^t ^^^ ^ ^s found to be 21" 3', or the current is 

a little over ^ period ahead of the impressed E. M. F. Ans. 

3. The E. M. F. E furnished by the alternator is equal to the 
current X impedance ; hence. 



^= /|/7?«-+-(2 7r«Z- ^^j^-^y = 535 volts Ans. 



Digitized by VjOOQIC 



§17 



ALTERNATING CURRENTS 



15 



In connection with this example, the student should note 
that while the E. M. F. furnished by the alternator is only 
535 volts, the pressure acrpss the terminals of the condenser 
is 663.5 volts. 



^# 



15, Resistance, Self-induction, and Capacity In 
Parallel. — In this case, shown in Fig. 18, the main cur- 
rent/is the resultant 
of the three cur- 
rents /„ /„ and /,. 
The current /, in the 
non-inductive resist- 
ance is in phase with fio. 18 

* E 
the E. M. F. and equal to ■^. Current /, will be equal to 

E 






^-KtlL 



and is 90** behind E in phase. 




Current /, is equal to 
%nn C E and is 90*" 
ahead of E in phase. 
The three currents 
will then be related 
as shown in Fig. 19. 
^^ is the difference 
between oc and oby 
and is found by sub- 
tracting /, from /„ 
as indicated by the 
dash line /,. Of 
course, if /, should 
happen to be greater 
than /„ the resultant 
current would be in 
the opposite direc- 
tion. This resultant oe combined with ofy which repre- 
sents /, to scale, gives o d^ which is the main current /. 
If the branches containing L and C also contained some 
resistance, /, and /, would not be at right angles to oa^ 
but would differ in phase by angles, the tangents of which 
are readily determined as explained in previous articles. 



PIO. 10 



Digitized by VjOOQ IC 



16 



ALTERNATING CURRENTS 



§17 



If such were the case, the diagram would become some- 
what as shown in Fig. 20 ; o e is the resultant of /, and /„ 
and od \s the main current / found by combining oe 
with of^ which represents /,. It is easily seen from these 




PlO. 90 

diagrams that the value of the main current, i. e., the 
length of od^ depends very largely on the phase relation 
of the currents in the branches. 

A striking example of the effect of the phase relation on the 
main current may be shown by the experiment illustrated in 
Fig. 21. Lamps /„ /„ and /, are alike ; /^ is connected in the 



]u n 




LffK 



U2L. 



Fig. 21 



main circuit, /, in the branch containing an adjustable induct- 
ance L, and /, in the branch containing the condenser C. 
The inductance L may be adjusted by sliding the coil over 



Digitized by VjOOQ IC 




§17 Alternating currents 17 

an iron core. If L is adjusted to the proper amount so that 
the curents in /, and /, are nearly the same in amount and 
are out of phase with the E. M. F. by nearly 90°, the main 
current in /, will be very much less than the currents in /, 
and /„ and lamp /, will burn at a dull red, while /, and /, are 
burning up to full brilliancy. 
This state of affairs could not 
occur with direct current flow- 
ing through a divided circuit. 
The diagram representing the 
three currents would be as 
shown in Fig. %%\ oc represents * 
the current in the condenser, 
and is nearly 90° ahead of the 
E. M. F. £"; the current in the in- 
ductance L is represented \^yob 
nearly 90° behind E in phase. 
The resultant current in lamp /, ^®* ** 

is represented by the line ^rf, and this resultant is very much 
smaller than either of the components o b or oc, 

16, The examples given in the preceding articles will 
serve to illustrate the composition and resolution of E. M. F.'s 
in such circuits as are commonly met with. The student 
should notice that in every case where such E. M. F.*s are 
combined or resolved, account must be taken not only of 
their magnitude, but also of their phase relation. For this 
reason such E. M. F. *s cannot be simply added together, as 
is done in dealing with direct currents, but the resultant 
sum must in all cases be obtained by using the polygon or 
parallelogram of forces. If these phase relations are kept 
in mind, many of the peculiarities in the behavior of the 
alternating current are easily understood. 

In working out problems, it is always best to draw out a 
diagram representing the different E. M. F.*s, as it makes 
the relation between them more easily understood. Quite 
a number of problems may be solved graphically by adopt- 
ing convenient scales for the different quantities and laying 



Digitized by VjOOQ IC 



18 ALTERNATING CURRENTS §17 

out the E. M. F. triangles. The resultant may then be 
scaled off the drawing and the result obtained more easily 
than by calculation. Examples of this method have been 
shown in connection with several of the preceding problems. 
Unfortunately, however, it is almost impossible to use the 
graphical method in a large number of cases arising in prac- 
tice, because the conditions are often such that the quanti- 
ties entering into the problem result in such long thin 
triangles and parallelograms that it is almost impossible to 
scale off any result accurately. In such cases the result- 
ant E. M. F. may be calculated by trigonometry from a 
knowledge of the sides and angles of the triangle or paral- 
lelogram in question. 



CAIiCUIiATION OF POWER EXPENDED IK 
AliTERNATING-CURBBNT CIRCUITS 

17, If a continuous current / flows through a wire of 
resistance R, the wire becomes heated, and the rate at which 
work is done in heating the wire is proportional to the 
square of the current / and to the resistance R; i. e., watts 

expended = /^ R. Since / = ■^, we have watts = £ I, 

Hence, it may be stated that, in a continuous-current 
circuit, if we wish to calculate the watts expended, we mul- 
tiply the current / by the E. M. F. £ necessary to force 




Curreni 



1ww^^AW^K5)fH|||||I||||p 



Fig. 23 



the current through the circuit. This is also true in a cir- 
cuit where the energy expended reappears in other forms 
besides heat. For example, we might have a direct-current 
dynamo i?, Fig. 23, sending current through a circuit ^J^/, 



Digitized by VjOOQIC . 



§17 



ALTERNATING CURRENTS 



19 



consisting of a resistance R, a motor J/, and a storage bat- 
tery B. The total power expended in the circuit from a to 
d will be the product of the current / and the E. M. F. £ 
across the circuit. Part of this energy = £, / will reappear 
as heat in the resistance R^ another part, equal to £, /, will 
reappear as work done by the motor M, and the energy 
expended in the battery, -£, /, will be stored up by virtue of 
the chemical reactions that are caused to take place by the 
current. 

18, If an alternating current be sent through a circuit, 
the power expended at each instant is given by the product 
of the instantaneous values of the current and E. M. F. It 
is seen at once, then, that the phase relation between the 
current and E. M. F. will have an important bearing on the 
power supplied, because the value of the E. M. F. corre- 
sponding to any particular value of the current will depend on 
their phase relation. In alternating-current circuits, there- 
fore, the power expended cannot usually be obtained by sim- 
ply taking the product of the volts and amperes as is done with 
direct currents. The effect of difference of phase between 
current and E. M. F. on the power expended can be well 




PIO. 94 

illustrated by means of the sine curves as shown in Figs. 24 
to 28 inclusive. Suppose an E. M. F. of maximum value E 
is in phase with a current of maximum value I, as shown in 
Fig. 24, the current being represented by the dot-and-dash 
curve, and the E. M. F. by the dotted curve. The power 



Digitized by VjOOQ IC 



20 ALTERNAWNG CtJRRENfS $ 1* 

at any instant, such as that represented by the point «, is 
proportional to the product of the ordinates ni and nfoi 
the r and e curves. If, therefore, an ordinate ng is erected 
at n proportional to this product, g will be a point on the 
power curve. In this way the power curve shown by the full 
line is constructed, and it shows the way in which the 
power supplied to the circuit varies with the E. M. F. and 
current. It should be noticed that in this case (current and 
E. M. F. in phase) the power curve lies wholly above the 
horizontal; that is, the work is all positive, or, in other 
words, power is being supplied to the circuit. This would 
be the condition if the current were flowing through a non- 
inductive resistance. 

19. Suppose, however, that the current lags behind the 
E. M. F. by an angle less than 90°, as shown in Fig. 25. 
The power curve is here constructed as before, but it is no 
longer wholly above the horizontal. The ordinate fg of 
the current curve is positive, while at the same instant that 
of the E. M. F. curve /r is negative; consequently, their 
product is negative, and the corresponding ordinate of the 



/- 



l/^rf -^'- 




Pio. 26 

power curve is below the horizontal. This means that dur- 
ing the intervals of time a' b' and c' d\ negative work is 
being performed ; or, in other words, the circuit, instead of 
having work done on it, is returning energy to the system 
to which it is connected. In Fig. 26 the angle of lag has 
become 90°, or the current is at right angles to the E. M. F. 
In this case the power curve lies as much above the axis as 



Digitized by VjOOQ IC 



8W 



ALTERNATING CURRENTS 



21 



below it, and the circuit returns as much energy as is 
expended in it. The total work done in such a case is 
therefore zero, and although a current is flowing, this cur- 
rent does not represent any energy expended. This would 
be nearly the case if an alternator were supplying current 
to a circuit having a small resistance and very large induct- 
ance, as in this instance the current would lag nearly 90^ 




behind the E. M. P. The primary current of a transformer 
working with its secondary on open circuit is a practical 
example of a current that represents very little energy. 
Such a current at right angles to the E. M. F. is, for the 
above reasons, known as a wattless current^ because the 
product of such a current by the E. M. F. does not represent 
any watt3 expended. 

20. Another example of a wattless current is that flow- 
ing into and out of a condenser when the resistance of the 
circuit is zero. If the angle of lag becomes greater than 90"*, 




PIO. »7 



the greater part of the work becomes negative, as shown in 
Fig. 27. If the angle of lag becomes 180°, as in Fig. 28, 



Digitized by VjOOQ IC 



ALTERNATING CURRENTS 



§17 



1. e., if the current and E. M. F. are in opposition, the work 
done is all negative, and, instead of the alternator doing 
work on the circuit to which it is connected, the circuit is 







PIO. 28 



returning energy to the alternator and running it as a motor. 
In the above diagrams the relation in phase between the 
current and E. M. F. is shown in each case by the lines oa 
and b^ respectively. 



21, The power curves in Figs. 25 to 28 show the instan^ 
taneous values of the watts expended in a circuit for differ- 
ent values of the angle of lag. What it is usually important 
to know, however, is the average rate at which energy is 
expended. Let o a^ Fig. 29, represent the effective value 

of the current /, which lags 
behind the effective E. M. F. 
£■ = ^ ^ by an angle 0. The 
average watts expended will be 
the E. M. F. E multiplied by 
that component of the current / 
which is in the same direction 
as the E. M. F. If a line per- 
pendicular to ^ ^ be drawn from a to d^ the line o d repre- 
sents the component of /, which is in the same direction 
as by and the watts expended are proportional to the prod- • 
uct b X o d. The same result would be obtained by 
multiplying the current t? a by the component of the E. M. F. 
in the same direction as the current, i. e., by the product 




Digitized by VjOOQ IC 



§17 ALTERNATING CURRENTS 23 

oa X oc. It IS usual to consider the current as resolved 
into two components, one at right angles to the E. M. F., 
and the other in the same direction, although it makes no 
difference in the numerical result which is taken. In Fig. 29 
od=^ oa cos <t>\ hence, watts = odxob^^oa cos V ob 
=^ E I cos 0. Or, oc = b cos 0, and watts — oc o a — ob 
cos ^ oa •=i E I cos 0. It may, therefore, be stated that the 
mean power supplied to an alternating-current circuity in 
watts, is equal to the effective volts multiplied by the effective 
amperes times the cosine of the angle by which they differ in 
phase. 

22. The fact that the product E I cos gives the watts 
delivered to a circuit may be proved as follows: 

Let b and o a. Fig. 30, represent the maximum values E 
and I of the E. M. F. and current, differing in phase by the 
angle 0. These lines are 
supposed to revolve uni- 
formly around o^ and the 
angle or, which is constantly 
increasing, always repre- 
sents the angular distance 
oi ob from the reference Fio. ao 

line ox. The angle remains constant; that is, ob Sindoa 
always keep a fixed distance apart. The instantaneous value 
of the E. M. F. is given by the expression r = E sin a. The 
current I also passes through a set of instantaneous values 
and lags behind E by the constant angle 0. The value of 
the current at any instant is given by the expression / = I 
sin {a — 0). The product of the instantaneous values ^ and i 
gives the instantaneous watts expended, or 

ei = El sin a sin {a — 0) 

Prom trigonometry, sin (a- — 0) = sin a cos — cos a sin 0;* 
hence, 

ei = El cos sin* a — E I cos a sin a sin 




♦This proposition may be proved as follows: Let A O B, Fig. 81, be 
the angle a and POB the angle <p; then, AOP = a — <;>. From any 



Digitized by VjOOQ IC 



24 ALTERNATING CURRENTS §17 

and the average value of the watts is 

av. ^ z = av. E I cos sin* a — av. El cos a sin a sin ^ 

or, since cos and sin are constant quantities, 

av. ^ / = El cos av. sin* or — E I sin av. sin a cos a 

The average value of sin a cos a is zero, since both pass 
through positive and negative values alike, and the average 
value of sin* a = ^ ; hence, 

El 

average ^ i = -^r- cos 

If the maximum values E and I are expressed in terms of 
their effective values, i. e., E = E^, I = I ^, we have 

average power = average ^/ = -f/cos (10) 



point P, draw PA perpendicular toOA, and P B perpendicular to O B. 

From B draw BC perpendicular to 
OA, and from P draw i'i? perpen- 




dicular to B C\ sin (a — ^) = 



0~P' 



but 



PA = DC\ and DC^ BC-BD\ 
hence, 

. , . BC-BD 
sin (or-^) = -Q^ 

BC B D 
~ OP OP 



Pig. 81 



Multiplying both numerator and 

T> Q 

denominator of the fraction ^-p by O B, and the numerator and 

B D 
denominator of -q^ by BP, we may write 

• / ^x BC^OB BD^BP 
sin(a-« = -qb^oTP^WP^O^ 

Now. ^^ = sin a, ^-^ = cos f because (9 ^ /* is a right angle. Also, 
the anirle DBP = a. because a4- <9^C = 1 right angle. zxAO BC 
^ DBP = 1 right angle; hence, -^-^ = cos or, and ^-^ = sin f so 
that we may write 

sin (a — ^) = sin a cos ^ — cos a sin f 



Digitized by VjOOQIC 



§17 ALTERNATING CURRENTS 26 



POWBB FACTOR OF A CIRCUIT 

23« If an alternating current of / amperes be flowing 
through a circuit, and the pressure across the terminals of 
the circuit is E volts, the watts that are apparently expended 
would be given by the product of the current and E. M. F., 
that is, by £ X /. The real watts expended are however 
obtained, as proved above, by the product oi E I and cos <t>. 

The ratio is called the power factor of the 

apparent watts ^ '' 

system. 

The i>OTver factor of a system may then be defined as that 
quantity by which the apparent watts expended in the sys- 
tem must be multiplied in order to give the true watts. From 
formula 10, it will be seen that the power factor is numer- 
ically equal to cos 0, and it is sometimes spoken of as the 
cos <t> of the system. 

24. If the current and E. M. P. are in phase, = and 
cos 0=1; consequently, the true watts expended under 
such circumstances may be obtained by simply taking the 
product Ex L When the angle of lag is 90'', cos = and 
the true watts expended is zero, i. e., the current is wattless. 
When becomes greater than 90^, cos becomes negative, 
thus showing that the circuit is delivering energy to the 
system to which it is connected. It will be seen, then, that 
it is quite possible to have large alternating currents flowing 
under high E. M. F.*s, and at the same time have very little 
energy expended. 

26, The power factor in the case of direct-current sys- 
tems is always unity ; but in cases where alternating current 
is used, it may vary from unity to zero. In most alternating- 
current systems the pressure is kept constant, or nearly so; 
hence, it follows that when a given amount of power is to be 
transmitted, the current will be smaller if the power factor 
is high than if it is low. This will, perhaps, be best illus- 
trated by means of an example. Suppose it is desired to 



Digitized by VjOOQ IC 



ALTERNATING CURRENTS 



§17 



transmit 100 kilowatts over a line by means of alternating 
current. The load on the line consists principally of motors, 
and will, therefore, be more or less inductive. We will sup- 
pose that the current lags behind the E. M. F. by an angle 
of 25*". Cos <l> is then equal to .90, and the power factor 
will be .90; hence, 

true watts = apparent watts X .90 
= volts X amperes X .90 
We will suppose that the pressure used in transmission is 
1,000 volts; then, 

100,000 = 1,000 X amperes X .90 

and the current necessary will be 111.1 amperes. If the 
power factor had been unity, only 100 amperes would have 
been required. This example will serve to show the neces- 
sity of having the power factor as high as possible. 



WATTIiBSS Ain> POWER COMPONENTS 

26, It was mentioned in connection with Fig. 29 that 
the current could be looked on as resolved into two compo- 
nents, one at right angles to the E. M. F. and the other in 
phase with it. This is shown in Fig. 32. The component 
at right angles to the E. M. F. is known as the Tvattless 



Povfer Component of I ^ j . jf .p. Jf 





Powe r Component of B 



Smtlv^ 



Pio. 88 



PI0.8S 



component of the current, and the part in phase is known 
as the power component. The E. M. F. may, in the same 
way, be looked on as divided into wattless and power com- 
ponents, as shown in Fig. 33. From these figures it is easily 
seen that the greater the angle of lag, the larger will be the 



Digitized by VjOOQ IC 



§17 ALTERNATING CURRENTS 27 

wattless component and the smaller the part that is really 
expending power in the circuit. 

27, Although wattless currents do not represent any 
great amount of power wasted, they are objectionable, 
because they load up the lines and alternators and thus limit 
their output as to current-carrying capacity. For exam- 
ple, an alternator might be furnishing a current of, say, 
20 amperes to a system having a very low power factor. The 
actual power delivered would be small, and the engine would 
not have to work hard to drive the dynamo. At the same 
time, the current of 20 amperes is circulating through the 
lines and the armature of the alternator, and thus will load 
up the lines and heat up the machine. As the current output 
of the armature is limited to a large extent by this heating, 
it is seen that the useful current that may be taken from 
the alternator is cut down by the presence of this wattless 
component. Alternating-current apparatus, such as induc- 
tion motors, etc. , are always designed so as to have as high 
a power factor as possible consir tent with economy. The 
use of condensers has been suggested for neutralizing the 
self-induction, thus increasing the power factor, and one 
manufacturing company has used condensers in connection 
with induction motors to cut down the lag in the current. 

Example. — An alternator generating an E. M. F. of 1,000 volts at a 
frequency of 60 cycles per second supplies current to a system of which 
the resistance is 100 ohms and the inductance .3 henry. Find the 
value of the current, angle of lag, power factor, apparent watts, and 
true watts. 

Solution. — We have 

E _ 1,000 



^ R} -V {^^tiLf 4/l00* + (2x3.14x60x.8)» 
1.000 



~ 150.8 
Reactance = 2 ir » Z = 118.04 ohms. 



= 6.68 amperes 



^ ^ 2nnL 118.04 ._ 
tan^ = -^- = -3^ = 1.18 



44—28 



Digitized by VjOOQIC 



28 ALTERNATING CURRENTS §17 

hence, ^ = angle of lag = 48* 80' ; power factor = cos ^ = .662; appar- 
ent watts - EI - 1,000 X 6.68 = 6,680; real watts = J?/ cos ^ 
= 1,000 X 6.68 X .662 = 4,389. Ans. 

The effect of the self-induction in the above example is, 
therefore, to cause a lag of 48° 30', and by so doing the cur- 
rent of 6.63 amperes is equivalent to only 4,389 watts trans- 
mitted; whereas, if there were no inductance and cos ^ = 1, 
this current would have been equivalent to 6,630 watts. 



TRANSMISSION UNES 

38. In transmitting power over lines by means of the 
electric current, a certain loss of energy always occurs, 
due to the resistance of the wire. This loss cannot be 
avoided, and all that can be done is to keep it down to 
within reasonable limits. Of course, the loss can be made 
as small as we please by increasing the size of the line con- 
ductor, but it is less expensive to allow a certain loss than 
to make the conductor very large. The lost energy in 
transmission lines varies greatly and depends largely on 
local conditions. Quite often it is about 5 or 10 per cent, 
of the power transmitted, and in some cases it is more than 
this, especially on long lines. The loss in the line results 
in a falling off in pressure between the station and the dis- 
tant end, the number of volts decrease, or drop^ as it is 
called, being obtained, in the case of continuous-current 
circuits, by multiplying the current by the resistance of 
the line. Evidently the drop will increase as the load or 
current increases, and if the pressure at the receiving end 
is to be kept constant at all loads, the pressure at the sta- 
tion must be increased as the current increases. 

29. The calculation of the size of wire to transmit a 
given amount of power over a given distance with a speci- 
fied loss is a simple matter in the case of a direct-current 
circuit, as it simply requires a wire of such a size that the 
resistance of the circuit shall not cause the loss to exceed 



Digitized by VjOOQ IC 



§17 ALTERNATING CURRENTS 29 

the specified amount. For example, suppose it is required 
to transmit 20 kilowatts a distance of 2 miles by means of 
direct current. The voltage at the delivery end is to be 
600, and the loss is not to exceed 10 per cent, of this, i. e., 
the allowable drop is 60 volts at full load. The full load 
current equals 

watts 

—^ = i|^^ = 40 amperes 

-R X / = 60 volts 

i? = 7^ = 1 26 ohms 
40 

The resistance of the whole length of wire from the station 
and back, or 4 miles, must not exceed 1.25 ohms, or the 
resistance per mile = .3125 ohm. By consulting a wire 
table, this is found to be about a No. 000 B. & S. wire. 

30. Self-induction of Xdne. — As long as alternating- 
current lines are not very long and the power is delivered to 
a load that is largely non-induct- 




the volts drop corresponding to a 
given size, by applying the same 
rules as used for direct current. 
However, if the lines are long and if the load is more or 
less inductive, as a load of induction motors, it is not safe to 
apply the direct-current methods. The reason for this is 
twofold. In the first place, a long line has considerable 
self-induction, and in the second, the effect of the self- 
induction of the line and the load is to throw the E. M. F. 
applied to the line out of phase with the current. Suppose 
A and B, Fig. 34, represent a cross-section of two line 
wires in which an alternating current is flowing. The cur- 
rent will set up magnetism between the wires, as indicated 
by the dotted circles, and as this magnetism is constantly 
changing with the changes in the current, an induced 



Digitized by VjOOQIC 



80 ALTERNATING CURRENTS §17 

E. M. F. IS set up that causes the drop in the line to be 
greater for a given current than it would if direct current 
were flowing. The farther the wires are apart, the greater 
will be the self-induction, because there will then be a 
larger area for the lines to thread between the wires. In 
cables, where the wires are insulated and twisted together, 
there is very little self-induction, but on overhead lines, 
where the wires must be spread several inches apart, it may 
have quite a decided effect on the drop. The self-induction 
of parallel copper wires may be calculated approximately 
from the following formula : 

L = .0805 + .74 log- (11) 

where L = inductance, per mile of wire, in millihenrys 
(thousandths of a henry) ; 
d = spread of wires (center to center) ; 
r = radius of wire. 
d and r may be expressed in any convenient unit so long 
as the same unit is used for each. 

Example. — (a) Find the inductance in millihenrys of 1 mile of 
line consisting of two No. 0000 B. & S. wires strung 24 inches apart. 
(tf) Find the reactance of this mile of line, assuming the frequency to 
be 60. (^r) Calculate the impedance of the line. 

Solution. — (a) In this case the diameter of the wire is .46 inch, 

or the radius is .23 inch. The value of ^ is 24 inches; hence, the 

24 
inductance of each of the wires will be Z = .0805 -H .74 log -^ = 1.575, 

and the inductance of the two lines will be 1.575 X 2 = 3.150 milli- 
henrys. Ans. 

(If) The reactance will be 2^«Z-, where L is the inductance in 
henrys. 8.15 millihenrys = .00315 henry; hence, reactance = 2 
X 8.14 X 60 X .00315 = 1.187 ohms. Ans^ 

(c) The impedance will be equal to j/A*" -i- (2 tt « L^. The resistance 
per mile of No. 0000 is .259 ohm, and the resistance of 2 miles would be 
.518 o htn. The im pedance of the 2 miles of line would, therefore, be 
-f/(T5i8)« + (l.l«7)-^ = 1.295. Ans. 

Table I shows the values of reactance and impedance of 
bare copper wires as given by Emmet. These values differ 



Digitized by VjOOQ IC 



§17 



ALTERNATING CURRENTS 



31 



§ 



£ 


£ 


Ri 


n3 o 




8 2 


S.^83 


III 


?. 




CB C 


ll 


« CO 


r> '^t 


»n o 


r>- 0» « in 


in 1 


li 


M M 


M M 


M M 


M M W « 


CO CO -4 


"" 1 


















«) 


^ S{ 














1 




1 


o « 


c*^ O 


ON W 


if> CO 


5- ? 5 § 


d in oo 


d ' 




as* 


« « 


« C*^ 


*? *? 


in in in 

M t-4 M 


^ 1 


Su JO 


















(A 


£5 


c^ m 


C- S 


O O 


8 8nR 2 


8 ?.?. 


8 ! 


1 o>- 


« c 


•H N 


« c^ 


Tt in 


r^ CO M in 


O "^ 


in 




5cS 


M M 


M M 


M l-l 


M l-l M « 


CO CO -^t 


in 
















g? 


— i> 


1 














«g 


tJ 2 


1^ o 


CO O 


oo M 


-^ t^ O c<> 


NO 00 M 


Tf 


"0 4« 


S c 


•-• w 


N C* 


W CO 


C«^ CO '<t '^t 


Tf '^f in 


m 


s.« 


S 


«« 


M M 


M M 


M M 


M M M M 


M M M 


M 


^s 


































■d 

a 


£ 

0/ 


P 


Is" 


0> u^ 


1^ 


2 8 85 


?,&8< 


R 




00 C 


•-. w 


O oo M '<t 


On »n to 


Tf 


1 


1" 


£-0 


M M 


M M 


1-1 M 


M M W W 


C« CO "^ 


in 


Is 


^ {? 


M M 


OO M 


?« ?,R 


in CO M 
to to <^ 


? 


0< 


& 


^s 


M M 


M M 


M M 




M M M 


M 


£ 


a 


25 


fl 


O* CO 


il 


^582 

w Tt CO « 


^8 a 


1 


o a 






M 


M M M CI 


CI CO •^ 


in 
















o 


•^s 




















o* 


M r>» 


CO CO 


CO CTi 


llga 


O CI M 


nO 


s 
i; 


'3 


ss 


M to 


«Jg 


to Tt O 

r-* t^ r-* 


1- 

r>. 


U 


































B'S 


s,s 


V> CO 


S'-§> 


vO OO 


w o 00 


8 8^8 


?> 


ii 


c c 


ll 


O vO 


t-* t-* 


CO q 


M ».; M ci 


r-* CO N 

CI to 4 


CO 

in 
















OCT 


•^SJ 
















S£ 


^1 


tJ ® 


1^ CO 


CO t^ 


^ o^ 


o o o o 


R tt fl 


c* 




xn m 


oo O 


o o 


s 


g.«« 


n 


aj«J 














•d 


















£ 


f^l 


o o 


« O 


» >oo 


?m 


2S 2 


o 


fl 


4) 


r>. M 


OO »o 


S 8 


CI 


« 


2° 


6§ 


m o 


O 1^ 


t^ CO w 


CO 


« 


c c 


^•o 






M 


M M M CI 


C« CO Tt 


in 
















w 


l-i 0) 
















cj 


=1 


o » 


00 <ri 


?;& 


o <8 


M vO O eo 


t^ d t-^ 


00 


s 




O « 


0» w CO 

in nO nO >0 


NO O nO 




fl< 


& 


.S- 


tr> m 


\n in 


u- in 


^. 


* 




> 


CN Tt 


w o 


»n o 


M CO O CO 


CO O O 


s 






tr> M 




in w 


Tt M in oo 


CO N OO 


s 




W CO 


'Tt xn 


o o 


O CO O o 


NO CO ►^ 


04 













M M M 04 


ci to -^i- 


xn 




§1* 




P 


8 « 


M « 


CO Tf »n vO 


t^ CO On 


O 




o n 

















Digitized by VjOOQ IC 



82 ALTERNATING CURRENTS §17 

slightly from those given by formula 11, owing to the num- 
ber of decimal places used in making the calculations. The 
difference is, however, not sufficient to be of practical 
importance. 

31, lAne Capacity. — The electrostatic capacity of over- 
head lines is usually of such small amount that it does not 
materially affect the drop unless the line is a very long one. 
In all except long transmission lines or long lines of under- 
ground cable, the line capacity may be neglected. The 
capacity per mile of two parallel wires stretched in air may 
be calculated by the following formula: 

C=:2i^ (18) 
log- 

where C = capacity in microfarads per mile of line; 

d = spread of wires; 

r = radius of wire, 
rfand r must be expressed in terms of the same unit. 

Example. — Find the capacity of 3 miles of line made up of two cables 
^ inch in diameter, strung 12 inches apart. 

Solution. — In this case </ = 12 inches; the diameter of the wire is 

iinch,orr =iinch; - = 48. and log 48 = 1.68124; hence, C= t^^t|t 

= .0115 microfarad per mile, and the total capacity is .0115 X 8 
= .0345 microfarad. Ans. 

32, Drop In Alternating-Current liines. — The 
amount of drop, i. e., the difference in voltage between the 
station end and the receiving end of the line, in an alter- 
nating-current circuit depends both on the impedance of the 
line and also on the kind of load that the current is supplied 
to. The relations are shown in Fig. 35. Let/ = cos ^ be 
the power factor of the load. Then the E. M. F. E' at the 
end of the line where the load is situated may be considered 
as made up of two parts, one of which t?^ is in phase with 
the current, and the other ^ ^ at right angles to the cur- 
rent. b represents the E. M. F. E' at the receiving end 



Digitized by VjOOQ IC 



§17 



ALTERNATING CURRENTS 



83 



of the line. The E. M. F. supplied by the generator must 
be the resultant of E and the E. M. F. necessary to over- 
come the impedance of the line. The E. M. F. necessary 
to overcome the line resistance will be in phase with the 
current and will be represented by o c. The E. M. F. neces- 
sary to overcome the line reactance will be represented by c d 
at right angles to the current line. Then,^rfis the E. M. F. 




Pio. as 

necessary to overcome the line impedance, and ^ ^ is the 
resultant of E' and od, and, therefore, represents the gene- 
rator E. M. F. £, The difference between E and E' is the 
drop in the line, and is represented to scale by 6/, which is 
obtained by striking an arc with o as center and ^ ^ as radius. 
For a given line, the resistance and reactance are easily cal- 
culated as explained above ; hence, the triangle ocd can be 
constructed. 

33. The required terminal E. M. F. E' is usually known, 
and the angle ^ is known from the nature of the load to be 
supplied. For example, if the load were all lights, the power 
factor would be very nearly 1 and the angle ^ could be made 
zero. If the load were all induction motors, the power 
factor, cos 0, might be about .85, corresponding to an 
angle of lag of a little over 31°. By laying out the dia- 
gram, to scale, the drop can be readily obtained. In case 
the problem is to obtain the size of wire required to limit 
the drop to a certain amount, the above diagram can be 
applied by using the cut-and-try method. A rough estimate 
of the size can first be made by treating it as a direct-cur- 
rent problem, and the size so determined applied to the 



Digitized by VjOOQIC 



ft4 ALtEfeNATlNG CtJkfeENtS git 

above diagram will give a general idea as to how much the 
allowable drop would be exceeded with alternating current. 
Two or three trials will usually be sufficient to show what 
size of wire will be necessary. From Fig. 35 it is easily 
seen that the drop in an alternating-current line cannot be 
obtained by multiplying the current by the impedance of 
the line, although if the load were very nearly non-inductive 
the product of current by impedance would give the line 
drop quite closely. In making calculations and in laying 
out diagrams as shown in Fig. 35, the triangles often become 
so long and thin that it is difficult to scale off results accu- 
rately, and it is convenient to use approximate formulas for 
estimating the line drop and determining the size of wire 
for a given transmission system. Formulas for making 
these calculations will be given later. 



ALTEBNATING-CTIRBENT MEASURING 
rt^STBUMENTS 

34. In measuring alternating E. M. F.'s and currents, 
we usually wish to know the square-root-of-mean-square, or 
effective, values, as these are used in most of the ordinary 
calculations. The maximum or instantaneous values are 
not used to any great extent. Ammeters and voltmeters 
for use on alternating-current circuits, as a general rule, 
therefore, indicate effective values, and most of such instru- 
ments will, if standardized by means of direct current, read 
effective values when connected to alternating-current 
circuits. 

There is not such a large variety of alternating-current 
instruments as of direct-current, since a large number of 
instruments adapted for direct-current work will not act at 
all with alternating current. Generally speaking, an instru- 
ment that will give indications with alternating current will 
work also with direct current, but the reverse is by no 
means true. Take, for example, the Weston direct-current 



Digitized by VjOOQ IC 



ill ALtERNAtiNG CURRENTS d6 

ammeters and voltmeters, which are widely used for direct- 
current measurements. The current flowing in a swinging 
coil reacts on a permanent field and thus produces a deflec- 
tion. If such an instrument were connected to an alterna- 
ting-current circuit, the coil would not move, because the 
current would be continually changing direction, and there 
would be as much tendency to turn one way as the other. 
These instruments are, therefore, not suitable for alterna- 
ting-current circuits, and should never be connected to 
them. This is true of any class of instruments where a 
deflection is produced by the current reacting on a constant 
magnetic field. 



CliASSES OF IK8TRUMBNTS 

36. For use in connection with alternating currents we 
are practically limited to four classes of instruments. 

1. Hot-wire ammeters and voltmeters. 

2. Plunger, or electromagnetic, ammeters and voltmeters. 

3. Electrodynamometers (ammeters, voltmeters, watt- 
meters). 

4. Electrostatic voltmeters. 



HOT-WIRE AMMETERS AND VOLTMETERS 

36, The first class of instruments depends for its action 
on the heating effect of the current. One type of hot-wire 
voltmeter is the Cardew. This instrument will indicate 
equally well on either direct or alternating current, and 
when it is calibrated with direct current, it will give the 
effective E. M. F. if connected to alternating-current mains. 

The hot-wire voltmeter may be used to measure current 
by connecting it across the terminals of a known non-induc- 
tive resistance, as shown at /?, Fig. 36. The voltmeter V 
in this case measures the drop from e to f through the resist- 
ance, this drop being equal to R I\ hence, if R is known, 
/ can be at once obtained, or the scale ot the instrument 



Digitized by VjOOQ IC 



36 



ALTERNATING CURRENTS 



§17 



might be so marked as to give the current directly. Hot- 
wire instruments may be shunted in this way because they 
possess very little self-induction. Alternating - current 
instruments that operate on the electromagnetic principle, 



• AMAAAAAAAAAr'' 




Pio. as 

and, therefore, have more or less self-induction, cannot be 
used in connection with shunts and give accurate results. 

37. Stanley Hot- Wire Instruments. — The most prom- 
inent example of the hot-wire type as used at present in 

America is the Stan- 
ley Instrument, 

manufactured under 
the patents of Hart- 
mann & B r a u n . 
These instruments 
are made for meas- 
uring either current 
or voltage, and when 
used to measure cur- 
rent, a shunt is em- 
ployed, as shown in 
Fig. 36. The am- 
meter and voltmeter 
have the same gen- 
^^^'^ eral appearance, 

shown in Fig. 37. The parts are mounted on a hard-rubber 
base held in a brass frame, so that the working parts are 



Digitized by VjOOQ IC 



§17 ALTERNATING CURRENTS 87 

thoroughly insulated from the case of the instrument. The 
wire that is heated by the current is up under the scale, 
and, hence, is not shown in Fig. 37. The knob n is for 
adjusting the pointer to the zero point in case the pointer 
should not return exactly to zero. The movements of the 
needle are damped by a small aluminum disk s that swings 
between the poles of the permanent magnet m^ thus making 




PlO. 88 

the instrument very dead beat. Fig. 38 shows the style of 
shunt used in connection with the ammeter; it consists of 
low resistance strips a connected to heavy terminals b. The 
ammeter is connected to the shunt by means of flexible 
cables furnished with the instrument. The cables are 
attached at cc, and it is important to see that the number 
on the cables corresponds to the number on the instrument, 
otherwise the indications will not be correct. In the amme- 
ter the hot wire is tapped at one or more points by means of 
flexible silver strips, and the sections of the wire are con- 
nected in parallel. This allows a long wire to be used, and 
yet makes the resistance so low that a small drop across 
the shunt is suflicient to operate the instrument. One of 
these silver strips is seen in Fig. 37, projecting below the 
central part of the scale. As already mentioned, an alter- 
nating-current ammeter cannot be shunted unless it has 
negligible self-induction. This type of instrument has prac- 
tically no self-induction, but an appreciable amount might 
be introduced by coiling up the cables leading to the instru- 
ment. If these cables are too long, they should on no 
account be shortened up in this manner; the slack should 
be taken up by doubling the cables back and forth on them- 
selves, not by winding up into a coil. 



Digitized by VjOOQIC 



ALTERNATING CURRENTS 



§17 



Fig. 39 shows the principle of operation of the Stanley 
hot-wire instruments, a bis the wire that is heated by the 
passage of the current. This wire is stretched tightly 
between two supports mounted on a base that expands and 
contracts, with changes in the room temperature, at the same 
rate as the wire. This compensates for changes in the tem- 
perature of the room in which the instrument is placed. 
Near the middle point is attached a second fine wire c, which 
is connected to the fixed post ^. To this second wire is 




PlO. 89 

attached a third wire d, which passes around the small 
pulley e and is held taut by the flat spring /. This wire is 
very fine and offers very little resistance to the bending 
action around the small pulley. The pointer A is attached to 
the vertical shaft that carries the pulley, and this shaft also 
carries the aluminum damping disk s; the shaft is mounted 
on jewel bearings, so as to turn with the minimum amount 
of friction. It will be noted that, by making use of the con- 
struction shown in the figure, a very small expansion of the 



Digitized by VjOOQ IC 



§17 ALTERNATING CURRENTS 39 

heated wire is sufficient to cause a movement of the needle 
over the scale. Since the wire is stretched tightly between 
the supports ad, a, very small expansion is sufficient to cause 
a considerable sag, and this in turn permits of a large deflec- 
tion of the pointer. The whole system being held taut by 
the spring, any slight expansion of ^^ is transmitted to the 
pointer and the movement is largely multiplied. The dotted 
outlines show, in an exaggerated way, the position taken up 
by the wires when a current is passed through ab. This 
method of causing the expansion of the wire to operate the 
needle is extremely simple, and at the same time gives great 
sensitiveness. Hot-wire instruments have the advantage 
that they can be used on circuits of any frequency and are 
equally well adapted to either direct or alternating current. 



PlilTNaER AITD MAGNETIC-VANE INSTRUMENTS 

38, Instruments of the plunder type have been used 
quite largely for ammeters and voltmeters in central stations, 
but they are being superseded by other types. In the old- 
style Westinghouse plunger ammeter, the conductor that 
carries the current forms a vertical coil within which hangs 
a straight core made of iron wire. This core is suspended 
from one arm of a balance, on the other arm of which is 
a weight that counterbalances the weight of the iron core, 
so that the pointer, which^ is rigidly fastened to the balance 
arm, will normally point to zero on the scale. As the strength 
of the current flowing through the coil varies, the pull on the 
plunger varies ; consequently the needle is deflected. 

The Westinghouse plunger voltmeter, which is similar in 
principle to the ammeter, has a coil with a large number of 
turns, in series with which is connected a high resistance in 
order to limit the current through the voltmeter. The 
so-called magrnetic-Tane Instniments are a modification of 
the plunger type. A small flat vane of iron, to which a 
pointer is attached, is mounted on an axis inside a coil. The 
tendency of the coil to rotate the vane is opposed by a spiral 
spring. 



Digitized by VjOOQ IC 



40 



ALTERNATING CURRENTS 



§17 



39, IncUned-Coll Instrumeiits. — Fig. 40 illustrates 
the principle of an alternating-current instrument that has 




FlO. 40 

been largely used for switchboard work in alternating-cur- 
rent plants. It is a modification of the magnetic vane type, 
and is known as the Tlioinson Incllned-coll instrument. 
It is made and used by the General Electric Company. A 
circular coil c^ shown in section, is mounted with its axis 
inclined to the horizontal. Through the center of the coil 
passes a vertical shaft that carries the pointer /. A small 
vane of iron v is mounted on the shaft at an angle, and the 
movement of the swinging system is controlled by the two 
flat spiral springs a^ a\ When a current flows through the 
coil, lines of force will thread it as shown by the arrows. 
The iron vane v will tend to turn so that it will lie parallel 
to these lines, as shown by the dotted lines t/, and in this way 
a reading is obtained. 
Inclined-coil amme- 
ters and voltmeters 
are made in a wide 
variety of styles suit- 
able for switchboard 
work and also for 
work where portable 
instruments are 
needed. Fig. 41 shows 
the external appear- 
ance of a horizontal-edgewise type of inclined-coil ammeter. 



PIO. 41 



Digitized by VjOOQIC 



817 ALTERNATING CURRENTS 41 

The edgewise type is now very largely used for switchboard 
work, as it occupies less space than the ordinary flat scale 
type and gives a form of scale that is very easily read. 
Fig. 42 shows the instrument with the cover removed; the 
dotted outline shows the location of the inclined coil. Com- 
mercial measuring instruments should, as far as possible, be 
dead beat, i. e., when deflected they should come to rest 
quickly. In order to dampen the movements of the pointer, 
a number of different methods are available. In the instru- 
ment shown in Fig. 42, the movements are dampened by an 



PlO. 43 

aluminum disk a that moves between the pole pieces of the 
permanent magnets ^, b. The disk is mounted on the lower 
part of the shaft that carries the iron vane. As the disk 
moves between the magnet poles, eddy currents are gener- 
ated in it, and the reaction of these currents on the mag- 
netic field produces a retardation that effectually dampens 
the movements of the needle. 



INDUCTION INSTRUMENTS 

40. The Westinghouse alternating-current ammeter, 
shown in Figs. 43 and 44, is claimed to be accurate for dif- 
ferent frequencies and for changes in temperature. It 



Digitized by VjOOQIC 



42 ALTERNATING CURRENTS §17 

consists of a laminated iron core k that has around one arm 
the actuating coil c through which the current flows. In 
parallel with this coil is a non-inductive resistance A, which 
is mounted where shown merely for convenience. The dif- 



PIO. 48 

ference of potential at the terminals of these coils will remain 
substantially constant for ail frequencies; hence as the fre- 
quency increases, the coil c will take less and the resistance h 
more current. The periphery of the disk /has the form of 
a spiral ; it projects into the air gap of the core k a maximum 
distance when the disk is at its zero position as a result of 
no current flowing through the coil c. 

Mounted on each arm of the core k are two short-circuited 
coils d^ e^ one side of each coil being located in a slot in the 
core, so that the field due to current set up in these coils is 
displaced with reference to the field set up by the coil r, and 
inasmuch as the currents in coils //, e are produced by 
induction from the current in coil r, there will also be a dis- 
placement of phase and a shifting magnetic field will result. 
The alternating field set up when current flows through the 
coil c induces a current in the disk /, which is acted upon 



Digitized by VjOOQ IC 



§17 ALTERNATING CURRENTS 43 

by a secondary alternating field differing in phase from the 
first and due to the induced currents in the coils d^ e by the 



PIO. 44 



original alternating field. The reaction of the current 
induced in the disk upon the secondary alternating field 



44—20 



Digitized by VjOOQIC 



44 ALTERNATING CURRENTS §17 

tends to rotate the disk ; this will be fully explained in 
connection with induction motors. As the current through 
coil c increases, the armature will be rotated; and as the 
radius of the disk becomes shorter, the degree to which the 
disk projects into the air gap will decrease, thus giving a 
more uniform scale than would be the case with a circular 
disk. The spiral spring n opposes the rotation of the disk, 
which is supported in jewel bearings and is very evenly 
balanced. A permanent magnet m makes the instrument 
dead-beat. The shunt resistance h should have a temperature 
coefficient at least as high as that of the disk, so that as the 
resistance of the disk increases with the temperature and 
the torque exerted upon it consequently tends to decrease, 
the shunt resistance will also become heated and will, 
therefore, take less current, thus forcing more current 
through the coil c to compensate for the effect of heat in 
the disk. 

This type of instrument may be arranged to serve as a 
voltmeter and also as a wattmeter by the use of current and 
potential coils, as explained in connection with wattmeters. 



EliECTBODTNAMOMETEBS 

41. The alternating-current instruments, which are 
perhaps the most widely used, belong to the third class. In 
the class known as electrodynamometers, the current in 
a swinging coi! is acted on by a magnetic field produced by 
a fixed coil. The field produced by the fixed coil changes 
with the changes in the current, the arrangement being 
somewhat similar to that used in the Weston direct-current 
instrument, except that the permanent magnet producing a 
constant field is replaced by a fixed coil producing an 
alternating field. The Siemens dynamometer has already 
been described in a previous section. This instrument is 
largely used for standardizing purposes in laboratories, but 
it is not well suited to commercial work, because it is not 
direct reading, and also because it is not portable. An 



Digitized by VjOOQ IC 



§17 ALTERNATING CURRENTS 45 

electrodynamometer, if calibrated, with direct current, will 
indicate square- root-of -mean-square or effective values of 
alternating current. 

42. For commercial work, it is necessary to have instru- 
ments that may be more readily worked with and handled 



PlO. 45 



than the Siemens electrodynamometer, and that will also 
have the advantage of being direct reading. The Weston 
alternatingr-cnrrent voltmeter is a good example of the 



Digitized by VjOOQIC 



46 



ALTERNATING CURRENTS 



§17 



dynamometer principle as applied to a portable instrument 
In this instrument the swinging coil is mounted between 
two fixed coils, and is carried on jewel bearings. The move- 
ments of the coil are counteracted by two small flat spiral 
springs, which also serve to carry the current into the coil; 
in fact, the whole construction is similar to that of the 
Weston direct-current instruments previously described, 
except that the magnetic field is produced by the two fixed 
coils instead of a permanent magnet. The instrument gives 
direct readings in volts by means of a pointer attached to 
the vertical pivot that carries the coiL 



^asm^ 



« 



L 



I 



^f^MMMJ??/M}J?>}»^I»J»»,»Wl»»Mm,j^JjmM7T. 




PlO. 46 



43. Wagrner Altematlngr-Current Voltmeter, — Pig. 45 
shows two types of alternating-current voltmeter for switch- 
boards made by the Wagner Electric Manufacturing Com- 
pany; {a) shows the type used when the voltmeter is 



Digitized by VjOOQ IC 



11 



ALTERNATING CURRENTS 



4* 



mounted wholly on the front of the board ; {b) shows the 
type where the working parts are in a case behind the 
board, the scale being illuminated from the rear by means 
of lamps. Fig. 46 shows the construction of the type 
shown in Fig. 45 {a) ; Fig. 47 shows the construction of the 




voltmeter shown in Fig. 45 (^). In each case A is the 
swinging coil, and B^ B the fixed coils. The swinging coil 
is delicately mounted on sapphire bearings, and its move- 
ment is counteracted by the small spiral springs ^, s. The 



Digitized by VjOOQ IC 



48 



ALTERNATING CURRENTS 



§17 



pointer arm/ carries the pointer that moves over the scale 5, 
which forms a portion of a circular arc. In addition to the 
pointer arm, there is an index arm / that can be set at any 
desired point on the scale by means of the knob w. This 
index indicates the point at which the voltage should be kept 
so that any movement of the pointer on either side of the 
index is at once noted. In these instruments the movements 
of the pointer are very effectually damped by means of a 
small aluminum vane or tube V that moves to and fro in 
a cup containing oil. In Fig. 45 (b) and Fig. 47 the lower- 
case m contains the high resistance that is inserted in series 
with the voltmeter to limit the current flowing through the 
coils. 

WATTMBTKBS 

44. In order to measure the power supplied in an alter- 
nating-current circuit, we must have an instrument that 
will indicate the real watts expended, i. e., one that will 
give deflections proportional to E I cos 0. Such an instru- 
ment is called a \vattineter, and has been mentioned in 
the des ription of instruments used for direct-current 
measurements. 

A wattmeter must average up all the instantaneous values 
of the product of current and E. M. F. ; consequently, it 
must be so arranged that its indications will be affected by 





PlO. 48 



both. The dynamometer can easily be adapted to this work 
by changing the winding and connections of the swinging 
coil. Consider a circuit a b^ Fig. 48, in which energy is 



Digitized by VjOOQ IC 



§17 ALTERNATING CURRENTS 49 

being expended, and suppose for the present that it is con- 
nected to a direct-current dynamo. The watts expended 
may, in this case, be easily obtained by connecting an 
ammeter C in series with the circuit, and a voltmeter K 
across it, so as to get the values of the current and E. M. F., 
the product of which gives the required watts. This method 
would not work, however, if the circuit were connected to an 
alternator y4, as shown in the figure, because it would take 
no account of the phase difference between current and 
E. M. F. In order to do this, it is necessary to combine 
the ammeter and voltmeter into one instrument. This is 
done by winding the fixed coil of a dynamometer with a few 
turns of heavy wire and connecting it in series with the cir- 
cuit, while the swinging coil is wound with a large number 
of turns of fine wire and connected across the circuit. It is 
usual to connect a non-inductive resistance r in series with 
the swinging coil, in order to limit the current flowing in it. 
Since the resistance of the swinging-coil circuit is constant, 
the current flowing through it will, at all instants, be pro- 
portional to the E. M. F. acting on the circuit a b. The cur- 
rent in the fixed coil will also be equal to the current flowing 
in the circuit ; hence, the torque action between the two 
coils will at each instant be proportional to the product e /, 
and the average torque action will be proportional to the 
average watts. Such an instrument will, therefore, indicate 
the true value of the watts expended, because it takes 
account of the phase difference between the current and 
E. M. P. 

45, Portable Wattmeter, — The Siemens wattmeter, 
like the dynamometer, is not direct reading, and is, there- 
fore, not as convenient for commercial work as the portable 
direct-reading types, such as the Weston. It is, however, 
the standard wattmeter, and the one that is used for cali- 
brating other instruments, because, like the dynamometer, 
there are a few parts about it to change or get out of order. 
Fig. 49 shows a Weston portable -wattmeter. This is 
constructed about the same as the voltmeter, except that 



Digitized by VjOOQ IC 



60 



ALTERNATING CURRENTS 



^11 




PIG. 40 



the fixed coils are composed of a few turns of heavy copper 
conductor that carry the current. The heavy binding 

posts a, b at the side 
of the case are the 
terminals of these 
current -coils, and the 
small binding posts r, d 
on the top connect 
with the swinging coil. 
In using wattmeters, 
care should be taken 
not to get the connec- 
tions mixed, because 
if the current coil 
should, by mistake, be 
connected across the 
circuit, the instrument 
would in all probability be burned out, as the resistance of 
this coil is very low and the resulting current would be 
enormous. In order that the readings of a wattmeter may 
be reliable, the self-induction of the swinging coil should be 
very small. This is especially necessary if the instrument 
is to be used on a number of circuits having different fre- 
quencies. If the self-induction is high, the instrument will 
not read correctly for any other frequency than the one 
with which it was calibrated. The bad effect of thb self- 
induction of the movable coil can, however, be made of 
negligible amount by using a very high non-inductive resist- 
ance r, Fig. 48, in series with the coil. If this is done, the 
impedance of the fine-wire circuit becomes so nearly equal 
to the resistance that the current in the swinging coil 
becomes so nearly nn phase with the E. M. F. that the read- 
ings of the wattmeter are practically correct. 

46. Compensated Wattmeter. — In Fig. 48 it will be 
seen that the current in C is the sum of the currents in ^^ 
and V. A wattmeter with its current and potential coils 
connected as shown would indicate a number of watts in 



Digitized by VjOOQ IC 



§17 



ALTERNATING CURRENTS 



61 



^vvyv^^ 



excess of the number expended in ab^ because the effect of 
the series-coil is larger than it would be if the current in the 
circuit ab alone were passing through it. Of course, the 
current in r and V is very small, and if the current in ^ ^ is 
at all large, the error so introduced does not amount to 
much ; but if the current in ^ ^ is small, the error might be 
appreciable. In the Weston compensated ^w^attmeter, 
the wire connecting to the swinging coil circuit is laid along- 
side the turns of the coarse-wire coils, and is so connected 
that the current in the fine wire flows in a direction opposite 
to that in the coarse wire. The excess of current in the 
coarse-wire coil is, 
therefore, offset by the - • - -^^^ 

counter-magnetizing 
action of the fine-wire 
turns, and the instru- 
m e n t indicates the 
watts expended in the 
circuit or device to 
which the wattmeter 
is connected. 

Fig. 50 shows the 
connections of the in- 
strument. A and B 
are the current termi- 
nals connected to the 
current coils c^ c\ The 
compensating coih^ is 
connected in series 
with the swinging 
coil D^ and the protec- 
tive resistance R, The 
potential binding 
posts that are ordina- 
rily used for measuring the power supplied to a given load 
are a b. When a reading is taken, the button k is pressed, 
thus allowing current to pass through the swinging coil. 
A third binding post / is provided for use when the field 




Digitized by VjOOQ IC 



62 ALTERNATING CURRENTS §17 

and pressure terminals are connected to independent circuits. 
Such connections are required when the instrument is being 
checked by passing a current through the current coils and 
applying a variable pressure to the potential coil; also, in 
cases where a test is being made with a constant current 
and varying pressure. In case an independent potential 
circuit is used in this way, the potential terminals are 
connected to posts /^, thus cutting out the compensating 
coil. The small resistance r takes the place of coil e^ so that 
the resistance of the potential circuit remains unaltered. 
The wattmeter shown in Pig. 49 is not provided with a 
compensating coil; hence, the middle binding post / is not 
provided. Wattmeters are made in a large variety of forms 
both for switchboard and portable work, and can be obtained 
to cover almost any desired range. 

47. Thomson Incllned-Coll Indlcatini^r Wattmeter. 

Fig. 51 shows a Thomson inclined-coil, horizontal-edgewise, 
indicating wattmeter as used by the General Electric Com- 
pany on alternating-current switchboards. Its construction 
is practically the same as the inclined-coil instrument shown 



FIO. 61 

in Fig. 42 except that the small iron vane is replaced by a 
coil of fine wire mounted at an angle to the shaft. The 
fixed coil A carries the main current, and is wound with a 
few turns of copper strip. This coil connects to the current 
studs B. The potential circuit connects to the upper 



Digitized by VjOOQ IC 



§17 ALTERNATING CURRENTS 63 

studs Cy and current is carried into the swinging coil by 
means of small spiral springs in the usual manner. The 
protective resistance in series with the swinging coil is 
usually mounted separately on the back of the switchboard. 

48. Wagner Indicating: Wattmeter. — The Wagner 
indicating wattmeters are constructed in the same manner 
as the voltmeters shown in Figs. 46 and 47. The only dif- 
ference is that the stationary coils are provided with a few 
turns of heavy conductor and are connected in series with 
the circuit, while the swinging coil, in series with its resist- 
ance, is connected across the circuit. In* the voltmeter, the 
armature and field coils are each wound with fine wire, and 
are connected in series with each other. 

Wattmeters of the electrodynamometer type are made in 
many different forms, but they all involve the same principle, 
though the disposition of the current and potential coils 
may vary in different instruments. In the Stanley watt- 
meters the coils are made approximately spherical in shape. 
In proportioning the coils and fixing their disposition as 
regards each other, the object aimed at is to secure as even 
a scale as possible, i. e., to prevent the divisions from being 
crowded together at either end of the scale. 

49. ' Sometimes it is necessary to know the total amount 
of energy expended in a circuit during a given interval of 
time, as, for example, in measuring the output of a station 
or the energy supplied to a consumer. For this purpose it 
is necessary to use a recording wattmeter ^ i. e., an instru- 
ment that will record the number of watt-hours electrical 
energy supplied during a given period. One of the com- 
monest types of such an instrument is the Thomson record- 
ing wattmeter, already described. This is really a modified 
Siemens wattmeter, arranged so that the moving coil 
revolves so long as current is passing. Recording watt- 
meters are made in a large variety of styles and sizes, and 
information relating to their use will be given later in 
connection with the applications of alternating currents to 
power transmission. 



Digitized by VjOOQIC 



64 ALTERNATING CURRENTS § 1» 



BI^ECTROSTATIC VOLTMBTBRS 

60. Another class of voltmeter available for alternating- 
current work is that which depends on the repulsion or 
attraction of two surfaces carrying electrostatic charges. 
Such instruments have been used most largely in commer- 
cial work for measuring high voltages, but instruments 
are also made on this principle that are quite capable of 
measuring low voltages. One type used for measuring 



PIO. 53 

high potentials is that mvented by Lord Kelvin, and 
illustrated in Fig. 52. A set of fixed quadrants a^ a\ b, b' 
is mounted so that the aluminum vane vv' may swing 
between them on the pivot d. The fixed set of quadrants is 
connected to one side of the circuit and the swinging vane 
to the other, so that when they become charged, the vane is 
attracted and drawn in between the quadrants, and the 
voltage is indicated by the pointer. These voltmeters have 



Digitized by VjOOQ IC 



§17 



ALTERNATING CURRENTS 



65 



the advantage that they require no power whatever for 
their operation. This is sometimes of importance, especially 
when the instrument is connected to a high-potential circuit 
and left connected continuously. A very small current in 
such a case might represent a considerable loss of energy. 




Pig. 58 



51. Stanley Electrostatic Voltmeter. — Fig. 53 shows 
, type of electrostatic voltmeter used by the Stanley Electric 



Digitized by VjOOQ IC 



66 ALTERNATING CURRENTS §17 

Manufacturing Company on alternating-current switch- 
boards. It operates on the same principle as the voltmeter 
shown in Fig. 52, but the general arrangement is different. 
B^ B and C, C are fixed plates mounted on a hard-rubber 
base. These plates are protected by a hard-rubber cover- 
ing H to prevent leakage and also to obviate any danger 
of short-circuiting between the vanes. ^ is a movable 
aluminum vane, to which is attached the pointer, the move- 
ment of which is counterbalanced by the spiral spring 5. 
The fixed plates B^ B and the movable vane A are con- 
nected together and form one pole of the instrument. The 
fixed plates C, C are connected together and form the other 
pole. When the voltmeter is connected to the circuit, 
B and A being charged alike will repel each other, while 
at the same time C and A will attract each other, with 
the result that the vane is deflected an amount depending 
on the pressure of the circuit. Two plug receptacles 7", T 
are provided on the instrument, in addition to the regular 
terminals, so that it may be compared at any time with 
a standard instrument. The movement of the needle is 
damped or steadied by the vanes F moving in the partially 
closed boxes D, 

Other types of electrostatic instruments are made, but 
they all work on about the same principle. The electro- 
static ground detector, described later in connection with 
switchboard appliances, is constructed in practically the 
same manner as the electrostatic voltmeter. 



Digitized by VjOOQIC 



ALTERNATORS 



SINGLE-PHASE ALTERNATOBS 



GKNTEBAIi CHARACTERISTICS 

!• Dynamo-electric machines used for the generation of 
alternating E. M. P.'s are known as alternators. It has 
already been shown that 
the E. M. F. generated in 
the armature of a direct- 
current dynamo is essen- 
tially alternating, and that 
the commutator is supplied 
to change the connections 
of the external circuit so 
that the current in it 
may be direct. It fol- 
lows, therefore, that if the 
proper terminals of a con- 
tinuous-current armature 
were connected to two 
collector rings in place of 
a commutator, the current 4 V 

furnished would be alter- pio. 1 

nating. In the majority of cases, however, alternator 
armatures are not wound in the same way as those for 
direct current. Consider a horseshoe electromagnet as 
shown in Fig. 1. When such a magnet is excited by means 

§18 

For notice of oopyrigbt. tee page immedUtely following Uie title page. 




Digitized by VjOOQ IC 



2 ALTERNATORS § 18 

of the coils on its two limbs, lines of force will flow out 
of the north pole N into the south pole S, as indicated 
by the arrows. The two pole faces are shown in the lower 
diagram, and the rectangular coil of wire C is supposed to 
be moved across the pole face N to the position shown by 
the dotted outline in front of 5. When the coil is in the 
position shown under the north pole, a small movement of 
the coil to the right will not cause a very large change in 
the number of lines threading it; consequently, only a small 
E. M. F. will be induced. While the conductors are moving 
under the pole pieces, the E. M. F. will be practically 
uniform if the field is uniform, and when the coil has reached 
the position shown by the dotted line, the E. M. F. will 
again be zero. The E. M. F. has, therefore, passed through 
one alternation, or half cycle, while the coil has been moved 
through the distance a b. This E. M. F. curve may be of 
the shape shown in Fig. 2, the portion at y being fairly 
uniform while the conductors are moving 
under the poles; or it may have a differ- 
ent shape, depending on the shape of 
the coil and pole pieces as well as on the 
Pio. s ^j^y jj^ which the magnetic lines are 

distributed. No matter what the shape of the curve ay b 
may be, the E. M. F. passes through one alternation when the 
coil is moved a distance equal to that from the center of one 
pole to the center of the next. If the coil C be moved back 
from 5 to N^ the same set of values of the E. M. F. is gener- 
ated in the opposite direction; hence, by moving the coil 
from N to S and back, the y 

E. M. F. passes through 
one complete cycle, as 
shown in Fig. 3. The 
arrangement shown in 
Fig. 1 would, therefore, 
constitute an elementary Pio. 8 

alternator, and the E. M. F. would be set up by movements 
of the coil back and forth across the pole faces, there being 
no rotation at all. Instead of moving the coil back and 






'■ 7 



Digitized by VjOOQ IC 



§18 



ALTERNATORS 



8 



forth, the same effect could be produced by moving the coil 
forwards continuously in front of a row of poles, as shown 
in Fig. 4. As the coil C moves past the poles, it cuts the 
lines of force first in one direction and then in the other, 
thus producing the alternating E. M. F. represented by the 
curve below. It should be noted that while the coil moves 
through the distance between one north pole and the next 
pole of the same polarity, the E. M. F. passes through one 
complete cycle. The distance from a to r, therefore, corre- 
sponds to 360° on the E. M. F. curve, and abX,o 180°. For 
every pair of poles passed, the E. M. F. passes through a 







VZ7 



Fig. 4 

complete cycle of values; hence, it follows that the number 
of cycles per second, or the frequency of an alternator, is 
equal to the number, of pairs of poles that the armature 
winding passes per second. If the number of poles on the 

machine is /, the number of pairs of poles is =^, and if the 

coil is moved past the poles s' times per second, the fre- 
quency n will be 






(1) 



44—30 



Digitized by VjOOQIC 



4 ALTERNATORS § 18 

3. Instead of the single coil (7, Fig. 4, being used by 
itself, three other coils, shown dotted, might be connected 
in series or parallel with (7, and the whole four moved together 
in front of the poles. If the coils were connected in series, 
it is evident that the total E. M. F. produced would be 
increased, because all the E. M. F.*s generated in the turns 
of the different coils would be added up. If they were con- 
nected in parallel, the E. M. F. would be the same as that 
produced by the single coil, but the current-carrying capacity 
would be increased, because there would now be four circuits 
to carry the current in place of one. It should be noted 
particularly that no matter how many coils there are, or 
how they are connected together, the frequency remains the 
same so long as the speed s and the number of poles is con- 
stant. In other words, the frequency of an alternator does 
not depend on the way in which the armature is wound. 
Connecting the coils in series is equivalent to making the 
winding of one coil of a large number of turns; connecting 
them in parallel amounts to the same thing as winding in 
one coil with a heavy conductor. As long, therefore, as the 
coils are all moved simultaneously, as is always the case, the 
frequency is not affected in any way by the scheme adopted 
for winding and connecting up the armature. 

3. It is evident that an alternating E. M. F. would be 
set up in the coil or set of coils. Fig. 4, if the magnet were 
moved and the coils held stationary. Also, both coils and 
magnet might be stationary and an E. M. F. still be induced 
by causing the lines of force threading the coils to vary. 
These three methods give rise to the following three classes 
of alternators: 

1. Those in which the armature coils are moved relative 
to the field magnet. 

2. Those in which the field is moved relative to a fixed 
armature. 

3. Those in which the magnetic flux passing through a 
fixed set of coils is made to vary by moving m^s§es of iron, 
called inductors, past them. 



Digitized by VjOOQ IC 



§18 



ALTERNATORS 



For convenience in referring to alternators, we will sup- 
pose that the armature is the revolving part and the field 
fixed, though it must be remembered that actual machines 
may be built with any of the three arrangements mentioned 
above. 



CONSTRUCTION OF ALTEBNATOBS 

4. The earlier type of alternator is that in which 
the coils are mounted on a drum and revolved in front of a 
magnet consisting of a number of radial poles. Alternator 




Fig. 5 



armatures may be of the ring, drum, or disk type, but the 
drum style is used almost exclusively in America. If we 
suppose the poles, Fig. 4, bent into ^ circle ^n<i the coils 



Digitized by VjOOQ IC 



6 ALTERNATORS § 18 

mounted on a drum revolving within the poles, we will have 
one of the earlier types of alternator. This arrangement 
is shown in Fig. 5, except that in this case the machine 
is provided with eight radial poles and eight coils on the arma- 
ture, giving a style of winding in common use for machines 
used on lighting circuits. In this case, there are as many 
coils on the armature as there are poles on the machine; 
but a winding might easily be used in which there would be 
only half as many coils as poles. There is a large variety of 
windings suitable for alternators, and the designer must 
select the one best suited to the work that the machine must 
do. In Fig. 5, the coils C are shown bedded in the slots/ 
on the circumference of the iron core P, which is built up 
of thin iron stampings. These coils are heavily taped and 
insulated and are secured in place by hardwood wedges w. 
This makes a style of armature not easily injured, and the 
use of the dovetailed slots and wooden wedges does away 
with the necessity of band wires. As the armature revolves, 
the coils sweep past the pole faces, and the E. M. F. is gen- 
erated in the same way as in Fig. 4; i. e., the movement of 
the coils relative to the pole pieces becomes one of transla- 
tion rather than of rotation. 

6. Alternators are generally required to furnish a high 
voltage, and, in consequence, the armature coils are usually 
connected in series. Care must be taken in connecting up 
such windings to see that the coils are so connected that 
none of the E. M. F.'s oppose one another. By laying out 
a diagram of the winding, the manner in which the coils 
must be connected will be easily seen. This has been done 
in Fig. 6, which shows diagram matically the winding of the 
armature in Fig. 5. The coils are represented by the heavy 
sector-shaped figures, and the connections between them by 
the lighter lines. The circles in the center represent the col- 
lector rings of the machine, and the radial lines that part of 
the coil which lies in the slot, that is, the part in which the 
E. M. F. is generated. The circular arcs joining the ends 
of the radial lines represent the ends of the coils that project 



Digitized by VjOOQ IC 



§18 



ALTERNATORS 



beyond the laminated armature core. The drawing is made 
to show the coils at the instant the conductors in the slots 
are opposite the centers of the pole pieces. At this instant 
the E. M. F. will be assumed to be at its maximum value 
and that the direction of rotation is such that the conduct- 
ors under the north poles have their E. M. F.*s directed 




from the back of the armature toward the front. These 
E. M. F.*s will be denoted by an arrowhead pointing toward 
the center of the circle, since the inner end of the radial lines 
represents the front or collector-ring end of the armature. 
The E. M. F.*s in the conductors under the south poles 
must be in the opposite direction, or pointing away from 
the center. After having marked the direction of these 
E. M. F. *s, it only remains to connect the coils so that the 



Digitized by VjOOQ IC 



8 ALTERNATORS § 18 



Digitized by VjOOQIC 



§18 ALTERNATORS 9 

current will flow in accordance with the arrows. Starting 
from the collector ring R, and passing through the coils in 
the direction of the arrows, it is seen that the connections of 
every other coil must be reversed; i. e., if i, 1\ 2^ 2\ etc., 
represent the terminals of the coils, 1' and 2' must be con- 
nected together, also 2 and 5, and so on. The end 8 is 
connected to the other collector ring and the winding thus 
completed. The connections of such a winding are quite 
simple ; but if not connected with regard to the direction of 
the E. M. F.'s, as showh above, the armature will fail to 
work properly. For example, if 1' were connected to 2, 2' 
to 3y and so on around the armature, the even-numbered 
coils would exactly counterbalance the odd-numbered ones, 
and no voltage would be obtained between the collector 
rings. Of course, in this case all the coils are supposed to 
be wound in the same direction, as is nearly always done in 
practice. The connections shown in the diagram, Fig. 6, 
are shown between the coils in Fig. 5. It should be noted, 
in passing, that this constitutes an open-circuit winding; 
that is, the winding is not closed on itself, like that of a 
continuous-current drum or ring armature. A large num- 
ber of alternator windings are of the open-circuit type, 
which is better adapted for the production of high voltages, 
because it admits of a large number of turns being connected 
in series. 

6. Most alternating-current dynamos of the revolving- 
armature and stationary-field type are built on much the 
same lines as direct-current multipolar machines. Usually, 
however, they have a larger number of poles. Fig. 7 shows 
a Fort Wayne (Wood) alternator with revolving armature 
and stationary field with inwardly projecting poles. This is 
an eight-pole machine with an armature winding similar to 
that shown in the diagram; Fig. 6, and represents a type of 
machine largely used for lighting work. The two collector 
rings r, r are seen mounted on the shaft inside the bearing, 
and are connected to the armature winding by heavily insu- 
lated leads «, n. The terminals of the machine are at /, /' 



Digitized by VjOOQ IC 



10 ALTERNATORS § 18 

and leads o, o^ connect to the shunt across the series field 
coils, as explained later. The commutator^ or rectifier^ is 
shown at m and is used to commutate the current passing 
around the series field coils. The exciter is in this case 
driven from the alternator shaft. Fig. 7 is typical of belt- 
driven revolving-armature alternators. In some makes, the 
commutator or collector rings are placed outside the bearing 
and the connections brought through a hole in the shaft. 
This allows the bearings to be placed closer together, but it 
makes the insulation more difficult and it is questionable 
whether there is any advantage in it. 

Fig. 8 shows a Westinghouse alternator with revolving 
armature designed for a low speed (150 revolutions per 
minute) in order to allow direct connection to a waterwheel. 
The machine is of 650 kilowatts capacity, and is provided 
with 4 collector rings because the armature has a two-phase 
winding. 

7. The number of poles on these machines is made large, 

in order to obtain the necessary frequency without running 

the machine at too high a speed. It is evident from what 

has been pointed out that for every revolution of the 

armature the E. M. F. passes through as many complete 

cycles as there are pairs of poles, and the frequency must 

P 
be « = ^ -y, where / was the number of poles and s the num- 

ber of revolutions of the armature per second. 
Since « = 4-^ (2) 

Ai 

or ^ = — (3) 

Therefore, with a given frequency «, the number of poles 
must' be made large if the speed s is to be kept down. For 
example, if an alternator has 8 poles and runs at a speed of 
900 revolutions per minute, its frequency will be f X ^^ 
= 60 cycles per second. If we attempted to obtain a 



Digitized by VjOOQ IC 



§ 18 ALTERNATORS 11 

frequency of 60, which is a very common one, by using a 
two-pole machine, the speed would have to be ^ = ^j^, or 
60 revolutions per second, or 3,600 revolutions per minute, 
a speed altogether too high for a machine of any size. 



Fig. 8 

8. It follows from the above that if the frequency is fixed, 
as it usually is, and it is. desired to run an alternator at a 
given speed, it must be made with such a number of poles/ 

that the condition n =^ ^s will be fulfilled. Alternators 

are often required to run at a specified speed in cases where 
they are to be coupled directly to waterwheels or engines. 



Digitized by VjOOQ IC 



12 



ALTERNATORS 



§18 



This leads to the designing of a large number of special 
machines suited to these conditions, because alternators, on 
account of the relation that must be preserved between fre- 
quency, speed, and number of poles, cannot be adapted to 
different conditions of speed and voltage by changing the 
armature winding, as is done with direct-current machinery. 
The number of poles used on conlmercial machines varies 
greatly, as there is a wide range of frequencies and speeds 
to be met. Alternators are built with the number of poles 
varying all the way from 4 up to 60 or 80 and sometimes 
more. The number of poles usually increases with the size 
of the machine, because the speed of the larger dynamos is 
necessarily less than that of the smaller. For example, one 
Westinghouse 1,200-kilowatt alternator designed for direct 
connection to a steam engine has 40 poles and runs at a speed 
of 180 revolutions per minute, thus delivering current at 
60 cycles. The number of poles, the output, and speed of 
some smaller sized machines are given below. 

TABIiE I 
ALTERNATORS 



Number 
of Poles 


6o Cycle 


Number 
of Poles 


125 Cycle 


Output 
Kilowatt 


Speed 
R. P. M. 


Output 
Kilowatts 


Speed 
R.P.M. 


8 

12 

i6 


75 
150 
250 


900 
600 
450 


10 

14 
16 


30 
120 
200 


1,500 
1,070 

940 



9. The E. M. F. curve furnished by an alternator of the 
type shown in Fig. 7 would not follow the sine law. Such 
machines, with heavy coils embedded in slots, usually give a 
curve that is more or less peaked and ragged in outline, and 
are best adapted for lighting work. For purposes of power 
transmission, it is desirable to have a machine giving a 



Digitized by VjOOQ IC 



§18 



ALTERNATORS 



13 



a 

-B +-^- 



^PMek' 



-4 



Pio. » 



smooth E. M. F. wave, and this can be obtained by adopting 
the proper kind of winding for the armature. The advan- 
tages and disadvantages of the different windings will be 
taken up in connection with alternator design, attention 
being paid here to the principles governing the generation 
of the E. M. F. and the connecting up of the armature 
coils. 

lO. The distance ef^ Fig. 9, from the center of one pole 
piece to the center of the next is called the pitch of the alter- 
nator. The relation between the 
pitch and the width of pole face A 
varies in different makes of ma- 
chines, but in a large number of 
American alternators the dis- 
tance B between the poles is made 
equal to the width of pole face A 
or one-half of the pitch, and the 
pole pieces cover 50 per cent, of the armature. The shape 
of the E. M. F. curve is* determined largely by the rela- 
tive shape of the coils and pole pieces and the way in 
which the conductors are disposed on the surface of the 
armature. 

The width of the opening W in the coil 
should not, in general, be much less than the 
breadth of the pole piece A, It has been 
found that it may be slightly less without 
doing any harm; but if made too narrow, 
trouble is likely to arise owing to the 
E. M. F.'s induced in different conductors of 
the same coil being opposed to one another, 
thus cutting down the total E. M. F. 
generated. 

This will be seen by referring to Fig. 10, 

where a coil of three turns is shown with 

^*^- ^^ its width of opening W less than the polar 

width A. When the coil moves across the pole face 

in the direction of the arrow, the E. M. F.'s induced 




Digitized by VjOOQ IC 



U ALTERNATORS §18 

in the two conductors a and jb will be in the same direc- 
tion, because they both cut lines of force in the same 
way. The consequence is that these two E. M. F.'s oppose 
each other, as will be readily seen by following the arrow- 
heads. When an alternator is loaded, the armature reaction 
causes the magnetism to crowd more or less toward one 
side of the poles, thus practically reducing the width of the 
magnetic flux, and on account of this it has been found 
possible to make the width W a little less than A without 
bad results. Usually, however, the width of the opening 
is nearly equal to that of the pole face. 



CAIiCTJX.ATION OF E. M. F. GEKBRATBB BY ALTBRXATORS 

11. It has been shown that the effective E. M. F. 
induced in a coil, when a magnetic flux ^ is made to vary 
through it according to the sine law, is 

c. 4.44 * Tn 
-^= 10" 

This is the case in an alternator producing a sine E. M. F. 
The flux 4>, which is caused to vary through the coils by 
the motion of the armature, is the number of lines flowing 
from one pole piece; T is the total number of turns con- 
nected in series on the armature ; and n is the frequency. 
This formula may be easily proved by remembering that 
the average volts equals the average number of lines of 
force cut per second divided by 10'. 

Let s = revolutions per second; 

/ = number of poles; 

9 = number of lines flowing from one pple, 
T = number of turns in series; 
2 7" = number of conductors in series. 

Each conductor cuts an average oi p 9 lines per revolu- 
tion, or p ^ s lines per second ; hence, 



Digitized by VjOOQIC 



§ 18 ALTERNATORS 16 

average E. M. F. = ^ — —. = — j^ 

But pxs = 2n 



hence, average E. M. F. = 



10» 



The effective E. M. F. is 1.11 times the average; there- 
fore, 

^ _ ^0Tnx 1.11 _ 4.44 Tn . 

or, the effective E. M. F. generated by an alternator is 
equal to 4.44 times the product of the number of lines flow- 
ing from one pole, the number of turns connected in series, 
and the frequency, divided by 10\ 

12. Formula 4 gives the effective E. M. F. at the col- 
lector rings when the alternator is run without any load, i. e., 
on open circuit. If the machine be loaded, the E. M. F. at 
the terminals will fall off from the value given by the above 
equation. This formula gives the total effective E. M. F. 
generated only when the turns T connected in series are so 
situated as to be simultaneously affected by the changes in 
the magnetic flux. This means that the conductors must 
be bunched together into heavy coils, like those shown in 
Fig. 5, if the maximum effect is to be obtained. If the 
winding were spread over the surface of the drum, as is 
done in direct-current armatures, the E. M. F. in one set 
of conductors would not rise to its maximum value until 
after the E. M. F. in the preceding set. For example, sup- 
pose we have an alternator wound with flat pancake coils, as 
shown in Fig. 11. The coil is here spread out to a certain 
extent, and the E. M. F.'s in the different turns will be 
slightly out of phase with each other, because they will not 
all come into and go out of action at the same instant. 
The total E. M. F. generated by such a coil will be the 



Digitized by VjOOQ IC 



16 



ALTERNATORS 



§18 



resultant sum of the E. M. F/s generated in the different 
turns, and the more these separate E. M. F.'s are thrown 
out of phase by spreading out the coil, the smaller will 
be the resultant terminal E. M. F. obtained. If the five 
turns of the coil shown in Fig. 11 were placed together in 
a slot, they would all be affected by the magnetic flux at 
practically the same instant. Hence, for a given length of 




PlO. 11 

active armature conductor, concentrated windings produce 
the maximum E. M. F. at no load, and if the winding is 
distributed, the terminal E. M. F. at no load is lowered. 
Both kinds of winding have their advantages and disadvan- 
tages, which will be taken up in connection with alternator 
design. For the present, the student will bear in mind 
that formula 4 gives the E M. F. when the machine 
is running on open circuit and when the winding is so con- 
centrated that all the conductors pass into and out of action 
simultaneously. 

13. The E. M. F. obtained at the terminals or collector 
rings of the alternator may be considerably less than that 
given by formula 4 when the machine is loaded, because a 
portion of the E. M. F. generated will be used in forcing 
the current through the armature against its resistance, 
and some of the E. M. F. will also be necessary to over- 
come the self-induction. In the ca§e of a direct-current 



Digitized by VjOOQ IC 



§ 18 ALTERNATORS 17 

dynamo, the pressure obtained at the brushes for any 
given load is equal to the total pressure generated less 
the pressure necessary to overcome the resistance of the 
armature. If / is the current and R the armature resist- 
ance, the lost volts are i?/, and the pressure at the brushes 
is Ef, = E — R /, where E is the total voltage generated. 
In the case of an alternator, the voltage at the terminals 
may fall off greatly as the load is increased, on account of 
the armature self-induction and also on account of the 
demagnetizing effect of the armature on the field, the 
falling off being much greater than that accounted for by 
the resistance. The effects of armature self-induction will 
best be imderstood by referring to Figs. 12 and 13. The 





PlO. 12 PlO. IS 

alternator is supposed to be run at a constant speed with a 
constant strength of field ; the total E. M. F. generated in 
the armature will therefore be constant, because the rate at 
which lines of force are cut does not change, no matter what 
current is taken from the machine. We will suppose that 
the alternator is working upon a non-inductive load, such as 
incandescent lamps, and we will represent the current in the 
external circuit by the line o x^ Fig. 12. The E. M. F. 
necessary to overcome the armature resistance will be repre- 
sented by ^ ^ in phase with the current o x and equal to RI\ 
the E. M. F. necessary to overcome the armature react- 
ance ''iT^nLI will be represented by ^^ 90° ahead of the 
current ; and the total E. M. F. necessary to overcome both 
resistance and reactance will be ^^ = I ^R^ + (^ ^ ^•^)'» 



Digitized by VjOOQIC 



18 ALTERNATORS § 18 

0° ahead of the current in phase. The resultant sum of this 
E. M. F. od Sitid the E. M. F. obtained at the terminals of 
the alternator must always be equal to the total E. M. F. 
generated B^ which is of fixed value so long as the speed 
and field strength remain constant. Since the alternator is 
working on a non-inductive load, the terminal voltage £' 
must be in phase with the current and in the same direction 
as the current line o x, Fig. 13. The line o d^ 0° ahead of ^ ;r 
and equal to o d, Fig. 12, represents the amount and direc- 
tion of the E. M. F. to overcome the armature impedance. 
Hence the total E. M. F. E must be the diagonal of a paral- 
lelogram that has its sides parallel to o d and o ;r, and of 
which d\s one side. The value of the terminal E. M. F. E' 
must therefore be of^ and it will be noticed that it is consid- 
erably less than the E. M. F. E, 

By examining these two diagrams, it will be seen that if 
the inductance of the armature is large, compared with the 
resistance, the line o d will be long and the angle nearly 
90**. Consequently, the terminal E. M. F. E' obtained from 
a given E. M. F. E will be very small. If sufficient current 
is taken from a machine with large armature self-induct- 
ance, the terminal E. M. F. may fall to zero; that is, all 
the voltage generated is used up in overcoming the impe- 
dance of the armature and practically a wattless current is 
flowing. In this case, the effects of armature self-induction 
only have been considered. It will be shown later that the 
armature may exert a powerful demagnetizing action on 
the field in case the machine supplies an inductive load 
and the apparent armature reactance when the machine is 
running may be much greater than the reactance measured 
at standstill. The predetermination of the falling off in 
voltage with increase in load is therefore a complicated 
matter, as it depends not only on the armature inductance 
but also on the various effects of armature reaction. The 
apparent reactance that the armature possesses due to the 
combination of these effects is often called the synolironous 
reactance to distinguish it from the ordinary reactance 
which takes into account the self-induction only. 



Digitized by VjOOQ IC 



§18 



ALTERNATORS 



19 



14. Alternators having high armature self-inductance 
may be short-circuited without much danger of burning them 
out. When a machine of 
this kind is short-cir- 
cuited the current does 
not rise to a very large 
amount, as with direct- 
current machines, 
because the voltage 
generated is required to 
overcome the inductance, ^'••« 
and is unable to set up a i— 
large current. As the 
load is increased, the 
E. M. F. falls off, at first 
slowly and then more 
rapidly, until, when a cer- 
tain current is reached, Av^pere: 
the terminal E. M. F. has fio. u 
dropped to zero, and no further increase in current can take 
place. This is illustrated by Fig. 14, which shows a curve 
taken from an alternator with an armature of fairly high 
self-induction. The normal full-load current of this machine 
is 25 amperes, and it is seen that as the load is increased, 
the terminal voltage keeps falling off, until at short circuit 
the current is about 47 amperes and the terminal voltage 
zero. Such a machine would probably not be injured by a 
short circuit, because it would be able to carry a current of 
47 amperes for some time without dangerously overheating 
the armature. At the same time such machines are not 
considered desirable, because good voltage regulation is of 
much more importance than incidental immunity from 
burn-outs. 




16. The student will see from the above that the output 
of an alternator may be limited if the armature self-induc- 
tion is too high, because the voltage may drop off before 
the machine is delivering the current that it is capable of 



44—31 



Digitized by VjOOQIC 



20 ALTERNATORS § 18 

doing without overheating. The output of alternators is, 
of course, aflfected by the heating of the armature con- 
ductors, just as in the case of direct-current machines, and 
the output is in some cases limited by this effect rather than 
by the self-induction. 



EXAMPLES FOB PRACTICE 

1. If an alternator is to run at 1,200 R. P. M. and to give a frequency 
of 60 cycles per second, how many poles must it have ? Ans. 6 

2. How many poles should a 60-cycle alternator have if it is desired 
to couple it directly to a waterwheel running 225 R. P. M. ? Ans. 82 

3. A ten-pole alternator runs at the rate of 1,500 R. P. M. The 
armature is provided with ten coils of 40 turns each, connected in 
series, and the flux through each pole is 1,000,000 lines. What E. M. P. 
will the machine give at the collector rings when running on open 
circuit? Ans. d,220Tolts 



FIEIiD EXCITATION OF AXTERXATOBS 

16. In most alternating-current systems, the voltage 
at the points where the current is distributed is kept con- 
stant, or nearly so. This means that the voltage at the 
terminals of the alternator must, as a rule, rise slightly as 
the load comes on, the amount of rise depending on the loss 
in the line. At any rate, the voltage at the terminals 
must not drop off, and, as it has been shown that, with con- 
stant field excitation, the voltage will fall off with the load, 
it becomes necessary to increase the strength of the field 
magnets as the current output of the machine increases. 
For accomplishing this there are two methods in use, which 
are analogous to those employed for the regulation of shunt- 
and compound-wound continuous-current machines. 

17. The simplest method is that indicated by Fig. 15. 
W^ represents the armature winding, the terminals 7", 7"' of 
which are connected to the collector rings Ry R\ connecting 
to the line by means of the brushes g^ k. The field is 
excited by a set of coils on the pole pieces represented by C, 



Digitized by VjOOQ IC 



§18 



ALTERNATORS 



21 



and current is supplied to these from a small continuous- 
current dynamo or exciter E, This is a small ^hunt-wound 
machine with an adjustable field rheostat r in its shunt field/. 
An adjustable rheostat R is placed also in the alternator 
field circuit. When the voltage drops, the fields may be 

WWWNAA/ 




PIO. 15 

Strengthened by adjusting the resistances R and r. This 
method, which is used with plain separately excited alter- 
nators, serves to keep the voltage right, and may be used 
with advantage when a number of alternators are supplied 
from one exciter ; but it is a hand method, and is therefore 
objectionable if theioad varies much. 

18. Another method, shown in Fig. 16, varies the excita- 
tion of the field in proportion to the current that the 
machine is supplying, and thus automatically keeps up the 
voltage. Each field coil in this case consists of two wind- 
ings similar to those used on compound-wound continuous- 
current dynamos. One set of windings is separately excited 
by means of the exciter £, and is provided with a rheostat R^ 



Digitized by VjOOQ IC 



23 



ALTERNATORS 



§18 



as in the previous case. The field of the exciter is also pro* 
vided with a rheostat r. The greater part of the. current 
furnished by the alternator flows through the series-winding, 
represented by the heavy coil S; and since this causes the 
magnetism to increase, the machine maintains its voltage. 




PIO. 16 

The separately excited coils set up the magnetism necessary 
for the generation of the voltage at no load, and the series- 
coils furnish the additional magnetism necessary to supply 
the voltage to overcome the armature impedance. 

19. The current flowing in these series-coils must not be 
alternating, because if it were it would tend to strengthen 
the poles one instant and weaken them the next, and on 
this account the current must be rectified before being sent 
around the field. This is accomplished by means of the 



Digitized by VjOOQ IC 




§ 18 ALTERNATORS 23 

commutator, or rectifler, C C\ which is mounted on the 
shaft alongside the collector rings. It consists of two cast- 
ings C, C (shown developed in the figure), 
which are fitted together and form a com- 
mutator of as many sections as there are 
poles in the machine. The alternate sec- 
tions are connected by the conductors Cy c\ 
as shown in Fig. 17, the light sections be- 
longing to one casting C and the dark to the ^ 
other C. Two brushes d and ^, which press ^' 
on the commutator, are so arranged that ^®* ^ 
one is always in contact with C, while the other touches O. 
The connections are as shown in the diagram. One ter- 
minal T of the armature winding connects directly to the 
ring R and thence to the line. The other terminal T' con- 
nects to one side of the rectifier C, and the other side C is 
connected to the remaining ring .-^'. By following the 
direction of the current, it will be seen that while the recti- 
fier causes the current to flow in the same direction in the 
series-coils 5, it still remains alternating in the line circuit. 
Take the instant when the coils occupy such a position that 
the current is flowing from the terminal 7", and mark the 
direction of flow in the different parts of the circuit by the 
closed arrowheads. The current will flow out on the line Z, 
back on Af to C\ through 5, flowing from left to right, back 
to C, and thence back to the armature. When the armature 
has turned through a distance equal to that between two 
poles, the current will be flowing in the opposite direction, 
as indicated by the open arrowheads; that is, it will be 
flowing out from T' to C, from C it will go to the brush e 
instead of d^ because it must be remembered that the recti- 
fier has turned through the same angle as the armature, 
and hence d has slid from C on to C\ From e the current 
flows through S in the same direction as before, back to C\ 
out on the line M, and back on L to T, The action of the 
rectifier is, briefly, to keep changing the connections of d 
and e as the current changes, thus keeping the current in 5 
in the same direction while it remains alternating in the 



Digitized by VjOOQ IC 



24 



ALTERNATORS 



§18 



line. Usually the brushes d^ e are placed on the commu- 
tator as shown by d and e\ Fig. 17, in order to have them 
farther apart, their action, however, being the same. A 
shunt resistance 5', Fig. 16, is usually placed across the 
coils 5, in order to adjust the compounding of the machine 
to the circuit on which it is to work, since by varying S\ the 
percentage of the total current passing around the field can 
be changed. This method of excitation has been largely 
used by the General Electric Company for their compound- 
wound alternators. 

20, Fig. 18 shows a method of compounding used on 
Westinghouse alternators. It is somewhat similar to the 



r- — HHwmi 




Fio. 18 



method last described so far as the action of the rectifier is 
concerned, but differs in the means used to supply the 



Digitized by VjOOQIC 



§ 18 ALTERNATORS 25 

series-fields with Current. The main current from the 
armature is not led through the series-coils, as in the last 
method, but is carried through the primary coil / of a small 
transformer that is mounted on the armature spider and 
revolves with it. In some machines, the laminated core ron 
which coils p and s are wound, is a portion of the armature 
spider, the armature disks being punched out so as to form 
spokes on which the coils are placed. The secondary coil 5 
is connected to the parts r, c^ of the rectifier. Coils/ and s 
are thoroughly insulated, so that it is easily seen that the 
series-coils have no electrical connection with the armature 
winding. The main current flowing through / sets up an 
alternating magnetism in the iron core r, and this in turn 
induces an E. M. F. in coil j, as explained later in connec- 
tion with transformers. As the armature current in / 
increases, the current in s and the series-coils also increases, 
and thus regulates the field excitation so as to keep up the 
pressure. The action of the rectifier is the same as before, 
its function being to make the current supplied from s flow 
through the series-field always in the same direction. One 
advantage of this method of compounding is that the high- 
tension current of the armature is not in any way connected 
to the field windings, thus rendering their insulation less 
difficult and reducing the liability of shocks to the dynamo 
tender. 



REVOIiVING-FIELiT> AND INDUCTOR 
AliTERNATORS 

21. It has been mentioned that it makes no difference in 
the case of an alternator whether the field or armature is the 
revolving part. It is hardly practicable to make a direct-cur- 
rent dynamo with a revolving field and stationary armature, 
because it is necessary that the brushes should always press 
on the commutator at certain neutral points that bear a 
fixed relation to the field, and the brushes would, therefore, 
have to revolve with it. This, of course, would be objec- 
tionable, because it is often necessary to get at the brushes 



Digitized by VjOOQ IC 



26 ALTERNATORS § 18 

while the machine is running. In an alternator, the brushes 
pressing on the collector rings do not have to bear any fixed 
relation to the field, consequently there is no objection to 
the use of a fixed armature, the current from which can be 
carried off by leads connected to the winding. Two col- 
lector rings are necessary for carrying the exciting current 
into the revolving field, so that the use of the stationary 
armature does not do away with moving contacts, but the 
collector rings for the exciting current are subjected to a low 
pressure compared with that generated in the armature, and 
hence are easy to insulate and handle. The revolving-field 
type has an advantage in that the armature, being station- 
ary, is easy to insulate for high voltages. This construction 
also admits of the ready use of armatures of large diameter, 
thus rendering such machines particularly adapted to slow 
speeds. 

22. Alternators of the revolving-field type have come 
into extensive use, and some have been built generating 



Fio. 19 



pressures as high as 10,000 or 12,000 volts. The arrange- 
ment of the parts of this type of machine is usually similar 



Digitized by VjOOQIC 



§18 



ALTERNATORS 



27 



to that shown in Fig. 19. This shows a portion only of the 
stationary armature, which is external to the revolving 
field. The armature core is built up of a large number of 
sectional stampings C provided with slots on their inner 
periphery, and the whole core structure is clamped in a cast- 
iron yoke A by means of the flange B, The armature 
coils D are held in the slots by means of wooden wedges in 
much the same way as in the revolving-armature machines. 
The field structure is made up of a cast-steel rim G carried 
by the arms //, which terminate in a hub keyed to the shaft. 
Laminated pole pieces E are bolted to G by means of bolts F, 
and the field spools K are held on by means of flanges O, 
These coils are connected together, and the leads Z, M are 
connected- to two collector rings on the shaft by means of 
which the exciting current is supplied. 




PIO. 80 

Fig. 20 shows a revolving-field alternator of the belt- 
driven type. This is a single-phase machine. The ter- 
minals of the stationary armature are shown at J/, while A 
IT the revolving field. Fig. 21 shows the general construc- 
tion. G is the laminated armature core supported by the 
cast-iron framework L\ f^f are the armature coils; Z? is a 



Digitized by VjOOQ IC 



28 ALTERNATORS § 18 

steel ring into which the pole pieces are dovetailed at e and 
held by means of keys. The exciting coils d are made up of 
copper strip wound on edge. 



PlO. 91 

23. In the inductor type of alternator, the collector rings 
for supplying current to the field may be done away with, 
and a machine obtained that has no moving contacts what- 
ever. In this class of machine, a mass of iron, or inductor, 
with projecting poles is revolved past the stationary arma- 
ture coils. The magnetism is set up by a fixed coil 
encircling the inductor ; and as the iron part revolves, the 
magnetism sweeps over the face of the coils, thus causing 
an E. M. F. to be set up. 

Fig. 22 shows the principle of the Westinghouse inductor 
alternator. In this machine the circular iron frame E sup- 
ports the laminations /% which constitute the armature core. 
These are provided with slots in which the coils 6^ are placed. 
Inside the armature is the revolving inductor A^ pro- 
vided with the projections C built up of wrought-iron or 
steel laminations. The circular exciting coil D is stationary 



Digitized by VjOOQ IC 



§18 



ALTERNATORS 



29 



and encircles the inductor A, thus setting up a magnetic 
flux around the path indicated by the dotted line. The pro- 
jecting poles C are all, therefore, of the same polarity, and 
as they revolve, the magnetic flux sweeps over the coils. 
Although this arrangement does away with collector rings, 
the machines are not so easily constructed as other types, 
especially in the large sizes. The coil D becomes large and 




Fig. 22 

difficult to support m place, and would be hard to repair in 
case of breakdown. The collector rings supplying a low- 
tension current to revolving-field coils should give little or 
no trouble; so, taking all things into consideration, it is 
questionable whether the inductor machine possesses much 
advantage over those with the revolving field. The Warren 
machine operates on the same principle as the Westinghouse 
machine shown in Fig. 19. 

34. The most prominent example of the inductor alter- 
nator as used in America is the Stanley machine. This is 
made in several different sizes, one of the larger machines 
being represented by Fig. 23. In this view, the halves A^ A 
of the stationary armature are shown drawn back, so as to 



Digitized by VjOOQ IC 



30 



ALTERNATORS 



§18 



allow access to the coils. When the machine is in opera- 
tion, these halves are bolted together by means of bolts 




passing through the lugs a^ a. The machine is double, 
there being two laminated cores r, c^ in the slots of which 
the coils d^ e are mounted (see also Fig. 24). It will be 



Digitized by VjOOQ IC 



§ 18 ALTERNATORS 31 

noticed that the coils marked d are placed midway between 
those marked e^ one set of coils overlapping the other. 
The result of this arrangement is that when half the con- 
ductors of one set of coils are directly under the poles, the 
conductors of the other set are out from under the poles; 
hence, when the current in one set is at its maximum value, 
the current in the other set is at its minimum value, thus 
making this particular machine deliver two currents that 
differ in phase by 90°. The machines can also be built to 
supply single-phase or three-phase currents, if desired. The 
revolving inductor is shown at /, Fig. 23, surrounded by the 
magnetizing coil M. All the polar projections / on one side 
of the coil are of the same polarity, and there is a similar 
set of opposite polarity on the other side of the coil. The 
construction will be understood by referring to Fig. 24, 



PlO. 94 

which shows an end view and section of a large Stanley 
machine, the different parts being lettered to correspond 
with those shown in Fig. 23. The heavy iron bars/, /serve 
to hold the core together and also to carry the magnetism. 
The path of the magnetic flux is indicated by the dotted line 



Digitized by VjOOQ IC 



32 ALTERNATORS § 18 

l-^S-i, and when the inductor revolves, the lines of force 
sweep across the stationary coils and thus set up the 
required E. M. F. Inductor machines have the advantage 
of having no moving wire about them, but machines of the 
revolving-field type, such as shown in Fig. 20, have the 
advantage of using small field coils that are easily repaired 
or replaced in case of accident. The revolving-field type is 
also cheaper to construct, especially in machines for low 
speed and frequency, and gives better voltage regulation. 



POLYPHASE ALTERNATORS 

36, The alternators discussed so far have all been con- 
sidered as machines that furnish one current only, and are, 
consequently, known as sing^le-pliase alternators. Men- 
tion has been made of machines which, being provided with 
two or more distinct sets of windings on their armatures, 
are capable of furnishing two or more currents to the lines. 
Such are known as polypliase, or multipliase, alternators. 
The two kinds in common use are two-phase, or quarter- 
phase, alternators and three-phase alternators. 

Two-phase, or quarter-phase, machines deliver two cur- 
rents that differ in phase by 90**. 

Three-phase machines deliver three currents that differ in 
phase by 120°. 



TWO-PHASE AliTEHNATORS 

26, Since a two-phase machine delivers two currents 
differing in phase by 90°, it follows that the two windings 
on its armature must be so arranged that when one set is 
delivering its maximum E. M. F., the E. *M. F. of the other 
is passing through zero. It has been shown that while the 
coils move from a point opposite the center of one pole piece 
to a point opposite the next of the same polarity the E. M. F. 
passes through one complete cycle; hence, if the E. M. F.'s 
generated by the two sets of coils are to be displaced 90% or 



Digitized by VjOOQ IC 



5 18 



ALTERNATORS 



33 



J- cycle, with reference to each other, it follows that one set 
of coils must be placed one-half the pitch behind the other. 
This brings one set of conductors under the poles while the 
other set is midway between them. 

27, It has been shown that for the most effective gen- 
eration of E. M. F. the wire on an alternator armature 
does not cover all the surface, and an armature such as 
shown in Fig. 5 could have another set of coils added, so 
as to produce an E. M. F. at 90° with that generated by 
the coils already shown on the drum. Such a winding is 
shown in Fig. 25, except that in this case there are only 




Pig. 25 



four coils in each phase instead of eight. This gives a 
common type of two-phase winding, and is used instead 
of the eight-coil arrangement, in order not to make the 



Digitized by VjOOQ IC 



84 



ALTERNATORS 



§18 



drawing too confused. Fig. 25, therefore, represents a 
two-phase winding having one group of conductors or 
one-half a coil per pole per phase. One phase is made 
up of the four coils A, which are connected in series, 
and the terminals a^ a' brought out to the collector 
rings i, ^. The four coils B^ which make up the second 
phase, are also connected in series, and the terminals d, d' 
attached to the light collector rings S, 4- The angular dis- 
tance by which the center of set B is displaced from set A 
is equivalent to 90°, or ^ cycle, as indicated in the figure, the 
angular distance from N to N being equivalent to 360°, or 
one complete cycle. 

28. Fig. 25 shows a common method of connecting up 
two-phase windings; namely, the method employing two 
distinct Circuits and four collector rings. This may be 
shown diagrammatically, as in Fig. 26. The windings are 
here represented by coils 1 and 2 connected to the collector 
rings a, a' and *, d\ These windings have no electrical con- 




PIO. 98 

nection with each other and connect to two distinct circuits. 

In Fig. 26 the two armature windings are independent, but 

they might be connected at their middle points (where they 

cross each other in the figure) and thus form an interlinked 

system. If this were done, the pressure between lines a' b 

E 
or a b' would be -7=, where E is the pressure between ^ ^ or 

1/2 

a! Vy 1. e., the voltage generated in each phase. Later 



Digitized by VjOOQ IC 



§18 



ALTERNATORS 



35 



on another type of interlinked quarter-phase winding is 
described in connection with alternators having closed 
circuit armature windings. The terms two phase and 
quarter phase are commonly applied to any machine that 
delivers two currents differing in phase by one-quarter of 
a cycle. As we have just seen, such a machine may have 
its windings interlinked or independent, and some writers 
refer to the interlinked type as quarter phase and to the 
independent type as two phase, though such distinction is 
by no means general. 

189. Sometimes, instead of using two distinct circuits 
with four collector rings, a common return wire is employed, 
as indicated in Fig. 27. Here one end of each of the phases 




PlO. 27 



is joined to a common return wire, and only three collector 
rings are necessary. If E represent the E. M. F. generated 
per phase, the voltage between a b and b c will be Zf, while 
that between a c will be E ^. This will be understood by 
referring to Fig. 28, the E. M. F. between a and c being the 
resultant of the two equal E. M. F.'s E at right angles to 
each other. 

30, Fig. 29 is the winding diagram showing the method 
of connecting up the coils of the armature, Fig. 25. This 
winding differs from that shown in Fig. 6, in that the con- 
nections of every alternate coil do not have to be reversed. 
By marking the direction of the E. M. F.'s by arrowheads, 

44—32 



Digitized by VjOOQ IC 



86 



ALTERNATORS 



§18 



as before, it is readily seen that the terminal i' must be con- 
nected to ^, 2' to 3, and so on. The difference in the method 
of connecting the two windings is caused by there being 
only four coils per phase in Fig. 29, whereas there are eight 




Pig. 89 

in Fig. 6. The coils of the second phase are shown dotted, 
and the connections between them are made in exactly the 
same way as those of the first phase. 

31. For delivering heavy currents at low voltages, arma- 
tures are sometimes wound with copper bars. In such 
cases there is usually only one turn, or two bars, per coil, 
and such a bar winding is shown in Fig. 30. This is the 
equivalent of the coil arrangement shown in Fig. 29, the 



Digitized by VjOOQ IC 



§18 



ALTERNATORS 



37 



connections between the bars being such that the current 
flows in accordance with the arrows. Windings of this 
kind are used on machines for furnishing heavy currents 
necessary for electric smelting or any other purposes that 
call for a large current. 




FIO. 80 



32, The two-phase windings shown in Figs. 25, 29, 
and 30 are of the simpler kind known as concentratecl or 
tinl-coll ^^ndlngs. The conductors on a two-phase arma- 
ture may be distributed in the same way as those of single- 
phase machines, there being two, three, or more coils per 
pole per phase. Both styles are in common use, and will 
be treated in greater detail in connection with alternator 
design. 



Digitized by VjOOQIC 



88 ALTERNATORS § 18 

33. Polyphase machines are used principally for operating 
power systems, though lamps are often run from them as well. 
For example, in Fig. 26 lamps could be connected across either 
of the phases, and in case a motor were to be connected, 
both phases would be used. The load on the different phases 
should be kept as nearly balanced as possible. It is usual to 
rate the output of alternators by the power they are capable 
of supplying a non-inductive circuit; that is, by the product 
of the volts and amperes that they can furnish without over* 
heating. The output of a two-phase machine is the sum of 
the outputs of the separate phases. For example, if it were 
said that a certain two-phase alternator had an output of 
150 kilowatts, at a voltage of 2,000, it would mean that the 
volts generated by each phase was 2,000, and hence the total 
full-load current with the machine working on a non-induc- 
tive resistance would be 75 amperes, or 37^ amperes per 
phase. Each line and the wire on the armature would there- 
fore have to be capable of carrying 37 J amperes. If the 
machine were working on an inductive load, the product of 
the volts and amperes would not give the output in watts, 
on account of the lagging of the current. The current in 
this case would have to be greater for a given output, and 
as the current output is limited by the size of the armature 
wire, it follows that an alternator will not deliver its full- 
load rating to an inductive circuit. If the above alternator 
were provided with only three lines and three collector rings, 
as in Fig. 27, the current in the common return wire would 
be 37^ X 4/2 = 53 amperes, nearly. The two outside wires 
would in this case be proportioned for 37^^ amperes and the 
middle wire for 53 amperes. 

Example. — A two-phase alternator is to have an output of 200 kilo- 
watts at a pressure of 2,000 volts, and is to be operated on a three-wire 
circuit. What will be the full-load current in each of the three wires, 
and what current must the wire on the armature be capable of carrying ? 

Solution.— Output per phase = 100 kilowatts. Hence, full-load 
current per phase = ^}JH^ = 50 amperes. The current in the two 
outside wires is therefore 50 amperes, and the wire in each set of arma- 
ture coils must be capable of carrying 50 amperes also. The current in 
the common return wire is 50 x f'S = 70. 7 amperes. Ans. 



Digitized by VjOOQ IC 



§ 18 ALTERNATORS 39 

The field magnet of polyphase machines is identical with 
that used for single-phase machines; in fact, the only dis- 
tinguishing feature of the former is the armature winding, 
the other parts of the machine being almost exactly the 
same, with perhaps a few minor changes, such as an 
increase in the number of collector rings, etc. 




THREE-PHASE AliTERNATORS 

34« The requirement of a three-phase armature winding 
is that it shall furnish three E. M. F/s diflEering in phase 
by 120°, or one-third of a complete 
cycle. This can be done by furnish- 
ing the armature with three sets of 
windings displaced 120° from each 
other. This means that phase No. 2 
must be one-third the angular dis- 
tance from one north pole to the next 
north pole behind phase No. 1, and 
also that phase No. 3 shall be dis- 
placed a similar angular distance behind No. 2. 
armature will deliver three E. M. F.'s differing in phase, as 
indicated in Fig. 31. The three E. M. F.'s are equal, and 
are represented by £"„ j5„ and £„ each being 120° behind 
the other. 

36. Fig. 32 shows a three-phase winding having one-half 
coil or one group of conductors per pole per phase. This is 
the three-phase winding corresponding to the two-phase 
arrangement shown in Fig. 25. The winding consists of 
three distinct sets of coils A, B^ and C, The angular dis- 
tance from the center of coil B to A \s equivalent to 120°, 
or is one-third of the distance from N to N\ also the coil C 
is displaced the same distance behind B, Each of these 
three sets is connected in series, leaving the three pairs of 
terminals a, a' ; b,b'\ c, c\ The coils are shown diagram- 
matically in Fig. 33, phase 1 being represented by the 



Digitized by VjOOQ IC 



40 ALTERNATORS § 18 

heavy lines, phase ^ by the dotted, and phase S by the light 
full lines. 

Fig. 34 shows the same winding as Fig. 32 with a some- 
what different mechanical arrangement of the coils. Coils 
belonging to the same phase are shaded and lettered alike 



■\ I 



\ "3 L' / 



/ • 7 ^ ' \ 



■7 



PIO. 88 



so that they can be readily distinguished. This arrange- 
ment of coils would be better in practice than that shown 
in Fig. 32, and is the one generally used for this type of 
winding. By allowing the straight coils to project at each 
end beyond the armature core, the crossings of the coils are 
much more easily arranged than in Fig. 32, thus making 
the winding easier to apply and insulate. 



Digitized by VjOOQ IC 




Pio. as 



\ 



\ I 



I 



PiaM 



Digitized by VjOOQIC 



42 



ALTERNATORS 



18 



STAR ANB DELTA CONNECTIONS 

36. There are two or three different ways in which the 
three pairs of terminals a-a'^ b-b\ c-c\ Fig. 32, may be con- 
nected to the collector rings. In the first place, each ter- 
minal might be run to a ring, as was done in the case of the 




PIO. 85 



two-phase armature. This would give three pairs of lines 
and six collector rings, as shown in Fig. 35. This is sel- 
dom, if ever, done in practice, as it complicates matters. 




PIO. 36 



and, moreover, it is not necessary. Again, one end of each 
of the three phases might be connected together, as shown 
in Fig. 36, and a common return wire d run from the com- 
mon connection, forming an arrangement similar to the 



Digitized by VjOOQ IC 



818 



ALTERNATORS 



43 



two-phase three-wire system, Fig. 27. This would necessi- 
tate four collector rings, and is used occasionally on alter- 
nators that are to be used considerably for lighting work. 
The return wire from the common junction is also some- 
times employed on three-phase distributing systems. It 
has been pointed out that the resultant sum of three equal 
currents displaced 120° is at all instants equal to zero. 
Consequently, if the resultant current is zero, there is no 
need of a return wire d^ so it may be omitted. However, in 
some cases, where power is distributed from transformers 
or three-wire systems, the different branches are apt to 
become unbalanced. Under such circumstances the com- 
mon return d is sometimes used. 

37. Omitting the common wire d of Fig. 36 gives the 
arrangement shown in Fig. 37, in which one end of each of 



J^M. 




Pio. 87 

the phases is joined to a common connection K and the 
other three ends are carried to three collector rings. This 
is a common method of connecting up three-phase arma- 
tures, known as the Y or star connection. The windings 
in Figs. 32 and 33 are shown connected up in this way. In 
connecting up the terminals a, a' \ b, b' \ c, c' to form a star 
winding, care must be taken to preserve the proper relation 
of the E. M. F.*s in the different sets of coils. When the 
E. M. F., in one set of coils is at its maximum, the E. M. F.'s 
in the other two sets are half as great and in the opposite 
direction. Suppose, then, that we take the instant when 



Digitized by VjOOQ IC 



44 ALTERNATORS § 18 

set A^ Fig. 33, is generating its maximum E. M. F. (con- 
ductors opposite centers of poles), and suppose that the 
E. M. F. in this set is directed away from the common junc- 
tion AT. Then terminal a will be connected to K. Since 
the current in the other two sets of coils is at the same 
instant one-half as great and in the opposite direction, they 
must be so connected that the current in them will be flow- 
ing toward the common junction. In order to satisfy this 
condition, the terminals d and c are connected to K. The 
remaining three terminals are connected to the collector 
rings i, ^, and S. 

38. Instead of connecting up the phases in the Y fashion, 
they may be formed into a closed circuit, as shown in 




f'in^' 



^^' 



PIO. 88 

Fig. 38, the collector rings being attached to the point 
where the phases join. This is known as the A (delta) or 
mesli connection. This method, like the last described, 
requires only three collector rings, and is extensively used. 
Fig. 39 shows the same winding as Fig. 33 connected 
up A, and it will be noticed that there is no common con- 
nection, as in Fig. 33. The three sets of coils are connected 
up in series, as before, leaving the three pairs of ter- 
minals a, a' ; b, b* ; r, c\ We will consider the currents in 
the coils at the instant when the current in set A is at its 
maximum ; that is, when the conductors are midway under 
the pole pieces. At this particular instant, the currents in 
the other two sets will be one-half as great and in such a 
direction that the sum of the currents taken around the 
closed circuit of the armature winding is zero. If the 



Digitized by VjOOQ IC 



§18 



ALTERNATORS 



45 



maximum current is represented by I, the value and direction 
of the currents in the three sets of coils must be as shown 




Pio. ao 

in Fig. 40. Starting from one end of phase A^ Fig. 39, by 
connecting a to the inner collector ring, we therefore pass 




.i£L 



Pig. 40 

through phase A in the direction of the arrows to the middle 
ring. From there we must pass through B against the 



Digitized by VjOOQIC 



46 



ALTERNATORS 



818 



arrows; hence the terminal b must be connected to the 
middle ring and the other end b' to the outer ring. From 
the outer ring the current must pass through C against the 
arrows; hence, the terminal c must be connected to the outer 
ring and the other terminal c' carried to the inside ring. 

39. Both of the above methods of connection are in com- 
mon use not only for alternators, but also for synchronous 
motors and induction motors; it is, therefore, important to 
bear in mind the methods of connection and the distinction 
between them. 



REIiATIOK BBTWBKN CURRENT, B. M. F., AND OUTPUT 

40, The E. M. F. and current output of a three-phase 
alternator depends on the scheme that is adopted for con- 
necting up the armature. Suppose the coils A^ B, and C, 
Fig. 41, represent the windings of an armature Y connected. 




PIO. 41 



Let « be the effective value of the voltage generated in each 
phase. The volts obtained between the lines a, b will be the 
resultant of the two voltages b \n A and B. Let E repre- 
sent the pressure between the lines. E must be equal to 
2 B cos 30°. 

Note.— The resultant of the E. M. F.*s in coils A and B, which 
is the line E. M. F. E, may be found as follows: Represent the 



Digitized by VjOOQ IC 



§ 18 ALTERNATORS 47 

E. M. F.*s in the three phases by the arrows a, b, c. Suppose the E. M. P. 
in -<4 to be directed away from the common junction ; then the E. M. F.*s 
in c and b will be directed toward the common junction. To add a and b 
we must draw b' from the extremity of a equal to and in the same 
direction as b. The resultant of a and b' will be d^ which is the line 
E. M. F. E, The resultant d is equal to 2 c cos SO** = c 4/3? 

In a three-phase H -connected alternator^ the voltage between 
any two collector rings is equal to the voltage generated per 
phase multiplied by \/S or 1. 7S2. 

Conversely : If the line voltage maintained by a three-phase 
y -connected alternator is E volts^ the voltage generated by 
each phase must be E divided by 4/^ or 1. 7S2, 

41. It is easily seen from Fig. 41 that if we have a cur- 
rent / flowing in any of the lines, the current in the phase 
to which it is connected must also be /. Hence, 

In a three-phase y -connected alternator ^ the current in the 
armature windings is the same as that in the line, 

42. The total output in watts will be the sum of the 
outputs of each of the three phases. The current in each 
of the three phases is /and the voltage is t\ hence, the total 
watts developed on a non-inductive load will be 3 f /. But 

f = -;z.. Therefore the total output W = ^-^ = i^ E I 
4/3 |/3 

where / is the current in the line, and E is the voltage 
between any pair of lines. Hence, 

The output, in watts^ of a three-phase ^-connected alterna- 
tor working on a non-inductive load is equal to 4/S or 1,7S2 
times the product of the line current and line E, M, F, 

43. For a Y-connected winding we may summarize the 
following formulas, in which 

B = volts generated per phase ; 
E = line voltage; 
W = total watts output; 
/ = line current; 
f = current per phase or current in windings. 



Digitized by VjOOQIC 



48 ALTERNATORS § 18 



E = 


2 £ cos 30° = e 4/3 


(5) 




E 


(6) 




■/=»•' 


(7) 




W=/^EI 


(8) 



44. In case the armature is delta connected, as shown in 
Fig. 42, the E. M. F. ^ generated in each phase is equal to 



j«iS/r 




the line E. M. F. £, because the different phases are con- 
nected directly across the lines. The current in the arma- 
ture windings, however, is not as great as that in the lines, 
because it divides at each of the collector rings. If /' repre- 
sents the current in each phase, the current in the lines will 
be 2 i' cos 30° = /' |/3 = 1. 732 /'. The total watts output will 

be 3 j' J? = 3 ^E = ^£ I=z 1.732 EI. Hence, it may be 

stated that 

In a three-phase delta-connected alternator^ the line volt- 
age E is equal to the voltage generated in each phase. 

In a three-phase delta-connected alternator^ the current I 
flowing in the line is equal to the current /' in each pliase mul- 
tiplied by i/S or 1. 782. 

Conversely : If the current flowing in the line is /, the 
current that the armature conductors must carry will be I 
divided by ^3 or 1. 782, 

The watts output of a three-phase delta-connected alterna- 
tor working on a non-inductive load is equal to 4/^ or i. 7S2 
times the product of the line current and line E, M. F. 



Digitized by VjOOQ IC 



§ 18 ALTERNATORS 49 

45. The following formulas relating to a delta winding 
may then be summarized: 



E=e 


(9) 


I=i'\^ 


(10) 


V3 


(11) 


W=EI'^ 


(12) 



46. It will be noticed that the expression for the watts 
output remains the same, whether the armature be con- 
nected Y or A. It follows, therefore, that the output of a 
three-phase armature is not altered by changing its connec- 
tions from Y to A, or the reverse. The Y method of con- 
nection gives a higher line voltage than the A for the same 
E. M. F. generated per phase, while the A connection cuts 
down the current in the armature conductors. The Y wind- 
ing is, therefore, best adapted for machines of high voltage 
and moderate current output, as it does not require such a 
high E. M. F. to be generated per phase. On the other 
hand, the A connection is more suitable for machines of 
large current output, as it keeps down the size of the 
armature conductors. The best style of winding for any 
given machine depends largely, therefore, on the work that 
it must do. These formulas regarding Y and A windings 
apply also to polyphase synchronous motors and induction 
motors described later. The formula W =. -f/ 4/3 gives 
the output only when the load is non-inductive, i. e. , power 
factor = 1, and balanced. If the load is inductive and 
balanced, the output is JF = 4/3 -£/ cos 0, where cos is 
the power factor of the load. 

Example 1. — A three-phase alternator has a capacity of 100 kilo- 
watts at a line pressure of 1,000 volts. What is the maximum line 
current that may flow in each of the three lines leading from the 
machine ? 

Solution. — We have, from formula 8, 

^= i^EI, or 100,000 = V8/l,000 



Digitized by VjOOQIC 



50 ALTERNATORS § 18 

- 100,000 ^„„ . 

hence, / = ;= = 57.7 amperes. Ans. 

1,000 X |/3 

Example 2. — (a) If the above armature be Y connected, what will 
be the current in the armature conductors ? (^) What must be the 
E. M. F. generated in each phase ? 

Solution.— (a) In a Y-connected machine, the current in the wind- 
ings must be the same as the line current, that is, 57.7 amperes. Ans. 

(d) From formula 6 we have 

£ 1.000 ^^ ,^ . 
fi = — = = — -= = 577 volts. Ans. 
4/8 |/8 

Example 8. — (a) If the same machine were changed to the delta con- 
nection, what would be the allowable maximum line current ? {d) What 
would be the line voltage with a delta-connected armature ? 

Solution. — (a) The winding is such as to allow 57.7 amperes in 
each phase ; hence, from formula 10, we have 

/ = /' |/3 = 57.7 |/3 = 100 amperes = line current. Ans. 

(^) The line voltage E would be equal to c, and would be 577 volts. 
The total output would be £"/ 4/8 = 577 X 100 X f^ = 100.000 watts, 
or the same as with a Y winding. Ans. 

47. It is seen, . then, in the above example; that by 
changing the Y winding over to A, the current output has 
been increased from 57.7 amperes to 100 amperes, while the 
line E. M. F. has been decreased, in the same proportion, 
from 1,000 volts to 577 volts, the total watts, however, 
remaining the same. 

48. The principal differences in the connections of mul- 
tiphase armatures, as distinguished from single-phase, have 
been given in the preceding articles; as far as the mechan- 
ical construction goes there is no essential difference. 
Toothed armature cores with the coils or conductors bedded 
in slots are now almost universal in alternators as well as 
direct-current machines. Of course, polyphase windings 
give rise to a larger number of coils to dispose of, and, 
therefore, usually give more crossings of the conductors 
or coils at the ends of the armature; but polyphase arma- 
tures are constructed essentially in the same manner as 
those used for single-phase machines. 



Digitized by VjOOQ IC 



§18 



ALTERNATORS 



SI 



MONOCYCIilC STSTBM 

49. The monocyclic alternator, brought out by Stein- 
metz, is intended for use in stations where the greater part 
of the load consists of electric lights, but where it is also 
desired to have a machine capable of operating motors as 
well. In cases where the motor load is large, it is usual to 
use a regular two- or three-phase system. The monocyclic 
system is now seldom installed. 

The monocyclic alternator is really a single-phase machine 
with a modified armature winding. The armature is pro- 
vided with a set of 
coils constituting the 
main winding, the ter- 
minals of which are 
connected to the two 
outside collector rings. 
Fig. 43. In addition 
to this winding, a sec- 
ond set of coils is pro- 
vided, which are placed 
on the armature 90** 
behind the main coils, 
in just the same way 
as shown for the two- 
phase machine. Fig. 25. 
This second set is un- 
like those in a regular two-phase machine in that the num- 
ber of turns in the teaser coils, as they are called, is only 
one-fourth that of the main set, and one end of the teaser 
set is attached to the middle of the main winding, instead 
of being brought out to a collector ring. The other end of 
the teaser winding is brought to the middle collector ring, 
as shown in Fig. 43. This second winding furnishes an 
E. M. F. displaced 90° from the main E. M. F., and of one- 
quarter its value, thus furnishing an out-of-phase pressure 
suitable for starting motors. If it is desired to run lights 

44-33 




PlO. 48 



Digitized by VjOOQ IC 



69 ALTERNATORS § 18 

only, the two outside wires alone are used, it being neces- 
sary to run the third wire only to places where motors are 
used. By referring to the figure, it will be seen that the 
E. M. F. between either of the o utside and the middle 
rings is equal to >/(J £y -\- {^Ey = .56 £, nearly. For 
example, if the main winding generated 1,000 volts, the 
pressure between the middle and outside rings ^ould be , 
660 volts, nearly. ^^^^ 

ALTERNATORS WITH CliOSED-CIRCUIT ARMA- 
TURE WINDINGS 

60. The windings for alternator armatures so far con- 
sidered have for the most part been of the open-circuit type. 
The A-connected three-phase armature is an exception, 
since the windings in this case form a closed circuit. Consider 
an ordinary two-pole direct- current armature winding. If, 
instead of connecting this winding to a commutator we con- 
nect two diametrically opposite points to collector rings, we 
will have a single-phase alternator with a closed-circuit dis- 
tributed winding having two paths in parallel. The current 
in the armature conductor would, therefore, be one-half the , 
current in the external circuit, as in a regular two-pole 
direct-current dynamo. 

If the closed circuit winding be tapped at four equidistant 
points, as shown in Fig. 44, and the terminals i', S\ and J^, ^', 
connected to collector rings, the current 
obtained from 1\ & will be at right 
angles to that obtained from ;^,' ^\ and 
a quarter-phase alternator will be the 
result. This style of winding has been 
used very extensively by the Westinghouse 
Company. If it is desired to operate a 
2' 4' two-phase three- wire system from such a 
Pio. 44 machine, three of the collector rings, or 

taps, are used ; as, for example, 1\ 3\ 2'. It is evident that 
no two of the taps could be connected together without 
short-circuiting a portion of the armature. If a three-wire 
two-phase system were operated from lines /, S\ fS^^ one 




Digitized by VjOOQ IC 



§18 



ALTERNATORS 



53 



phase would be between 1' and 2' and the other between S' 

and 2\ Also, if the E. M. F. between 1' and 3\ i. e., the 

E. M. F. per phase with the four-wire arrangement were E, 

then the E. M. F. between 1' and 2' or the E. M. F. in the 

E 

If a three-phase two-pole 



second case would be 



1.414' 

machine were required, the closed-circuit winding would be 
tapped at three equidistant points. For multipolar alter- 
nators with an ordinary multiple winding, there would be a 
tap to each ring for each pair of poles, as explained later in 
connection with rotary converters. With series- wound two- 
circuit multipolar armatures it would be necessary to have 
only one tap for each ring, the angular displacement of 
these taps depending on the number of poles and the num- 
ber of phases. 



3< Auxi/my 




Pig. 45 

61. Westlngliouse Two-Phase Compound-Woiind 
Alto^ktor. — Fig. 45 shows, diagrammatically, the con- 
nections for a Westinghouse two-phase or quarter-phase 



Digitized by VjOOQ IC 



54 



ALTERNATORS 



818 



alternator with closed -circuit armature winding. The dia- 
gram is drawn for a two-pole machine in order to simplify 
it as much as possible. The winding A is tapped at the 
four equidistant points i, 2, 5, -4, and these taps connected 
to the collector rings 1\ ^^ S\ 4'. In series with one tap of 
each phase are the two primary coils C, C of a revolving 
transformer mounted within the armature. The coils are 
wound on a core together with the secondary coil Z>, which 
is arranged so that the resultant magnetic flux produced 




'%m^ 




PIO. 46 

by C and C passes through D, The current induced in D 
is therefore proportional to the sum of the currents in the 
two phases. This current is rectified by means of the com- 
mutator £ (which would have as many segments as the 
machine has poles), and passed around the series-field, thus 
strengthening the field as the load is applied. The sepa- 
rately excited winding is not indicated in the figure, as 
this is supplied with current in the usual manner from a 
separate exciter. 



Digitized by VjOOQIC 



§18 



ALTERNATORS 



55 



52. Westlngrhorise Three-Phase Componiid-Woiind 
Alternator. — Fig. 4G shows the connections for a West- 
inghouse three-phase compound-wound alternator. The 
method of compounding is similar to that shown in Fig. 45, 
except that the series-transformer carried within the arma- 
ture has three primary 
coils C, C\ C*\ connec- 
ted in series with the leads 
as shown. The resultant 
flux induces a secondary 
current proportional to 
the sum of the loads on 
the different phases, and 
this current is passed 
through the auxiliary 
coils after being rectified. 
It should be noticed in Fig. 46 that the connections to one 
of the primaries are reversed. The reason of this is that 
the resultant flux produced by three currents OA^ OB^ OC^ 
Fig. 47, differing 120° in phase would be zero. If one of the 
coils is reversed, three fluxes O A^ B\ OC differing by 
60° are the result, and these three produce the resultant 
flux OF. 




(b) 



PlO. 47 



Digitized by VjOOQIC 



Digitized by VjOOQIC 



ALTERNATING-CURRENT 
APPARATUS 



TRANSFORMERS 

1. One of the principal reasons why direct current 
is giving place so largely to alternating is the ease with 
which the latter may be transmitted over long lines at 
high voltages, and then be transformed at the receiving 
end to currents of lower pressure suitable for operating 
lights, motors, or other devices. If power is to be trans- 
mitted over long distances by means of the electric cur- 
rent, it is absolutely necessary that high line pressures be 
employed, in order to make the cost of the conductors 
reasonably low. For a given amount of power transmit- 
ted, the current will be smaller the higher the pressure 
employed. The loss in the line, however, increases with 
the square of the current ; consequently, if the pressure on 
a line be doubled, it means that, with the same loss, only 
one-fourth the amount of copper will be required in the 
conductors. In other words, with a given amount of 
power to be transmitted and a given fixed amount of loss, 
the copper required will decrease as the square of the volt- 
age increases. By using high pressures, it is evident that 
a small line conductor may be made carry a large amount 
of power over a long distance, and still not have the loss 
any greater than if a low pressure and a very large and 
expensive conductor had been employed. 

§19 

For notice of copyright, see page immediately following the title pag^ 



Digitized by VjOOQ IC 



2 ALTERNATING-CURRENT APPARATUS §19 

2. Transmission with direct current at high pressure 
has never been used to any large extent in America,- 
though some plants of considerable size are in operation 
in Europe. This is principally because of the difficulty of 
building direct-current machines to generate the high 
E. M. F.'s necessary. The commutators on such high- 
tension machines are apt to give trouble, and, moreover, 
it is difficult to transform the high-tension direct currents 
at the other end of the line down to currents at pressures 
suitable for ordinary use. The alternating current is open 
to neither of these objections, because an alternator has no 
commutator to give trouble, and high-tension alternating 
currents may easily be transformed down. Devices used for 
changing an alternating current of one voltage to another 
of higher or lower voltage are known as transformers. 
Transformers may be used either to step-up the voltage, i.e., 
increase it, or they may be used to step-down^ or decrease, 
the line pressure. Whether the transformer be used to step 
up or down, the change in pressure is always accompanied 
by a corresponding change in the current, and the power 
delivered to the transformer is always a little greater thain 
that obtained from it. For example, suppose a current of 
20 amperes were supplied to a transformer from 1,000-volt 
mains. If the load on the transformer were non-inductive, 
the E. M. F. and current would be almost exactly in phase, 
and the watts supplied to the side connected to the mains 
(primary side of the transformer) would be 20 X 1,000, 
or 20,000 watts. The power obtained from the secondary 
side, or the side connected to the circuit in which the power 
is being used, would not be quite as much as this. Sup- 
pose the secondary E. M. F. were 100 volts; if there were 
no losses in the transformation, we would obtain 20,000 watts 
from the secondary, and the available secondary current 
would be *^g^ = 200 amperes. In other words, the 
decrease in E. M. F. has been accompanied by a corre- 
sponding increase in current. As a matter of fact, there is 
always some loss in conversion, and the secondary output is 
never quite equal to the power supplied to the primary. 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 3 

The ratio ^ -. — ^— gives the efficiency of the trans- 
watts input 

former. A good transformer is one of the most efficient 

pieces of apparatus known, some of large size delivering as 

much as 98.6 per cent, of the energy supplied. 

3. Transformers used for changing an alternating cur- 
rent at one pressure to another alternating current at 
another pressure are often called static transforiuers, 
because they have no moving parts. This is done to dis- 
tinguish them from rotary transformers, or rotary con- 
verters, which are used to transform alternating current 
to direct, or vice versa. Such machines always have 
moving parts, hence their name. 

4. Nearly all transformers are operated on constant- 
potential systems. The transformer is supplied with cur- 
rent from mains, the pressure between which is kept constant, 
and this current is transformed to one of higher or lower 
pressure, the secondary pressure also being constant or 
nearly so. 



THEORT OF THE TRANSFORMER 

5* A simple transformer is shown in Fig. 1. C is a 
laminated iron ring, on which are two coils P and 5. The 
coil Phas a number of turns Tp, and 5 has, we will suppose, 
a smaller number of turns T,. The coil P is the primary^ 
and is connected to the alternator mains across which the 
constant pressure Ej, is maintained. Coil 5, from which 
current is delivered, is called the secondary. We will suppose, 
for the present, that the resistance of both primary and 
secondary coils is negligible. The above is essentially the 
construction of the ordinary static transformer. It consists 
of two coils or sets of coils interlinked by an iron magnetic 
circuit. 

Suppose a voltmeter V be connected to the terminals of 
the secondary coil 5. The resistance of the voltmeter is 



Digitized by VjOOQ IC 



4 ALTERNATING-CURRENT APPARATUS § 19 

very high; consequently, a very small current will flow 
through the secondary, and we may, for all practical pur- 
poses, consider the secondary as on open circuit with no 
current flowing. The line E. M. F. Ej, will cause a current 
to flow through the primary coil, and this current will set 
up an alternating magnetic flux in the iron core. This 
alternating flux will set up a counter E. M. F. in the coil P^ 
which will be the equal and opposite of E^ since the coil is 



To Altematgr. 











/'" 


""*•* 


1 
1 

1 

1 






I 


I 


[1 ?y 
t lit 



\^ 









^ V 




'^.J 



Tv Alternator, 




Pio. 1 



supposed to have no resistance. If the maximum magnetic 
flux be and the number of cycles per second «, we will 
have 

« _ 4.44 ^ T^n 

^'^ ~ 10- 

The current that will flow in the primary when the sec- 
ondary is on open circuit is that current which is required 
to set up a magnetic flux ^ capable of producing a back 
E. M. F. equal and opposite to the applied E. M. F. Since 
the coil Pis provided with a closed iron circuit, it is evident 
that a very small current is able to set up a large magnetic 
flux; hence, the current required when the secondary is 
on open circuit may be, perhaps, only a fraction of an 
ampere. In other words, the applied E. M. F. is capable 



Digitized by VjOOQ IC 



§ 19 ALTERNATINO-CURRENT APPARATUS 6 

of forcing only a small current through the primary, on 
account of its high self-induction. 

6. The flux ^ set up by the primary coil also interlinks 
the secondary 5, and as it is continually alternating through 
it, an E. M. P. will be set up in the secondary coil, which 
will be 

4.44 ^T.n 



^.= 



W 



and %^T. <^) 

or E. = eJ^ (2) 

?• The ratio of the primary voltage to secondary, 
i. e., -7^, is called the ratio of transformation. It also 

' e: 

follows from the above that the ratio of transformation is 
equal to the primary turns divided by the secondary turns. 
For example, if a transformer be supplied with 1,000 volts 
primary and has 600 turns on its primary coil while there 
are 50 turns on the secondary, the ratio of transformation 
is 10, and the secondary voltage 1,000 X -^ = 100 volts. 
In this case the transformer reduces the voltage from 1,000 
to 100, but the operation could be reversed, that is, it could 
be supplied with 100 volts and the pressure raised to 1,000. 

8. It was assumed above that all the magnetic flux (P 
that threaded the primary coil also passed through the 
secondary, and in well-designed transformers this is nearly 
the case. However, some lines may leak across, as shown 
by the dotted lines ^, ^, r, ^, without passing through both 
coils. This is known as magrnetlc leakagre, and its effect 
on the action of the transformer will be noticed later. 

9. So far, in dealing with the action of the transformer, 
the secondary has been supposed to be on open circuit. It 
is now necessary to examine the transformer action when 



Digitized by VjOOQ IC 



6 ALTERNATING-CURRENT APPARATUS § 19 

the secondary is working on a load. While the construction 
of a transformer is exceedingly simple, the reactions that 
occur when it is loaded are by no means so simple, and 
for the sake of clearness we will first examine the action of 
an ideal or perfect transformer— one that has no resistance 
in its coils, no magnetic leakage, and no hysteresis or eddy- 
current loss in its core. Such a transformer would have an 
efficiency of 100 per cent., and after examining its workings 
we can easily note the effect of the introduction of one or 
more of the above defects, which are present to a greater 
or less extent in all commercial transformers. The fact 
that the efficiency of good transformers is commonly over 95 
or 96 per cent., and even rises to over 98 per cent., shows at 
once that the combined effect of all the above defects does 
not change the performance of the transformer very much 
from that of the ideal. 

10. In the first place, if the core reluctance were zero, 
the current, with open-circuit secondary, necessary to set up 
the magnetic flux would be infinitely small. All cores have, 
however, some reluctance; hence, the effect of reluctance 
in the core is to increase this no-load current, or maflr- 
netlzlngr current, as it is called. Since this magnetizing 
current is that which is caused to flow against the self- 
induction of the primary, it follows that it is a wattless cur- 
rent 90** behind the E. M. F. of the mains. It is important 
to keep the magnetizing current as small as possible; there- 
fore, the magnetic circuit should be made short and of 
ample cross-section. 



£, = £p ry^ in the secondary at no load. Now, if the 



11. We have seen that the primary induces an E. M. F. 
Is 

secondary circuit be closed, say through a non-inductive 
resistance made up of a number of incandescent lamps, a 
current will flow that will be in phase with the secondary 
E. M. F. When a transformer delivers current from its 
secondary, this secondary load current is directly opposite 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 7 

in phase to the load current in the primary. Action and 
reaction are always equal and opposite, and the counter 
secondary current represents the reaction against which the 
primary current works when the transformer is loaded. 
Whenever a current /, is taken from the secondary, a cor- 
responding current /^ flows in the primary. The total 
current in the primary will, therefore, be made up of two 
components, one of which is the magnetizing current m 
and the other 7^, due to the current /, in the secondary. 
The magnetizing power of the primary and secondary coils 
is proportional to their ampere-turns, that is, to 7^ /p and 
T^ /„ respectively. The load currents 7p and /, are opposed 
to each other ; hence, the total magnetizing effect of these 
two currents is 

Now the input Ep Ip is equal to the output E, /, for an 
ideal transformer, or 

E T 

Ip = I,-j^ = fg-jf- (3) 

-Cp Ip 

hence, we may write the total magnetizing effect of the two 
currents, 

/.^T,-/.T. = (4) 

That IS to say, when a current /, is taken from the 
secondary of an ideal transformer, the corresponding cur- 
rent Ip that flows in the primary over and above the 

£ 
magnetizing current m is /, ~, and the total magnetizing 

effect of the two currents is zero. This means that the 
flux in the iron core will remain constant, no matter what 
load is placed on the secondary. 

12, Since the magnetic flux is constant for all loads, it 
follows that the secondary induced E. M. F. will also be 
constant, and if the secondary coil has no resistance, the 
E. M. F. at its terminals, that is, the secondary line E. M. F., 



Digitized by VjOOQ IC 



8 ALTERNATING-CURRENT APPARATUS §19 

will not change as the load is applied. An ideal trans- 
former would, therefore, if supplied with a constant primary 
pressure, maintain the voltage between the secondary lines 
constant at all loads. This condition is approached quite 
closely in the best makes of modern transformers, the vari- 
ation in secondary voltage being not more than 1.6 to 
2 per cent., depending on the size. It is thus seen that 
while the transformer is most simple in construction, it 
adjusts itself exceedingly well to changes in load so as to 
maintain the desired constant secondary pressure, the whole 
automatic regulation being brought about by the interac- 
tions of the currents in the primary and secondary coils. 



-^jr 



ACTION OF IDEAIi TRAK8FOBMEB 

13, The action of a transformer without resistance, 
magnetic leakage, hysteresis, or eddy-current losses may be 

represented by Fig. 2, when 
^* working with an open-circuit 

secondary. Let O N repre- 
sent the magnetic flux; then 
the magnetizing current will 
be in phase with O N, and 
hence may be represented by 
O m. This current is 90** 
behind the primary impressed 
E. M. F. £p, and £^ will, 
therefore, be represented by 
the line O E^, 90° ahead of 
P«G- « Om. The secondary E. M. F. 

E, will be directly opposite in phase to E^, and will be 

T 
represented by 0E, = E^ -^. In this case the transformer 

takes the small current O m from the line, and since this 
current is at right angles to the E. M. F., it is wattless, 
cos = and £p m cos ^ = 0, and the transformer con- 
sumes no energy. If the reluctance of the magnetic circuit 



Digitized by VjOOQ IC 



§19. ALTERNATING-CURRENT APPARATUS 9 

were negligible, the magnetizing current O m would be 
very small. However, in these diagrams its length has 
been exaggerated with respect to the lines representing the 
other currents, in order to illustrate the action more 
clearly. 

14, When the secondary is connected to a non-inductive 
resistance, a current /. flows in the secondary, and the action 
of an ideal transformer in this case is represented by Fig. 3. 
The lines O N and O m represent the magnetic flux and the 
magnetizing current, as before. O E^ is drawn to scale to 
represent the primary E. M. F., and O E^ represents the 




"*9 



B» 




^1f 



PlO. 8 



PIO.4 



secondary E. M. F. Let R be the resistance of the non- 
inductive circuit to which the secondary is connected ; then 

E 
the secondary current will be /, = -^, since the resistance is 

non-inductive, and may be represented to scale by the line 

O din phase with O £,. The load current in the primary 

T 
corresponding to /, in the secondary will be 7p = /, -^ and 



Digitized by VjOOQ IC 



10 ALTERNATING-CURRENT APPARATUS § 19 

may be represented by the line O a opposite in phase to 
O d and representing I^ to the same scale that O d 
represents /,. In the case shown, E^ = \Ej,\ hence, O a 
must be one-half the length of O d. The total current flow- 
ing in the primary will be the resultant ol O a and O m, 
or O b = /p. This resultant current lags 0** behind the 
impressed E. M. F., and it is evident that the more the 
transformer is loaded, the smaller becomes, and the nearer 
the primary current gets into phase with the primary 
E. M. F. In other words, taking current from the second- 
ary acts as if it decreased the self-induction of the primary, 
thus allowing a larger current to flow. Since decreases as 
the load is applied, it follows that cos increases; hence, the 
power supplied to the primary increases. 

15. If the secondary of an ideal transformer furnishes 
current to an inductive load, such as motors, the secondary 
current /, lags behind the E. M. F. E, by an angle /? of 

O — |f 7" 

such amount that tan /3 = — -^ — , where L is the induct- 
ance of the circuit and R the resistance. This action is 
represented by Fig. 4. The magnetizing current O w, mag- 
netic flux O N, primary E. M. F. £p, and secondary E. M. F. 
E, are all represented.as before. The secondary current /„ 
however, lags behind E, by the angle /?; consequently, the 
corresponding primary load current O a is behind the pri- 
mary E. M. F. Ep by the same angle /?, because /^ is oppo- 
site in phase to O d. The total primary current, being the 
resultant of O m and O a, will be <9 ^, lagging 0° behind E^. 
It will be noticed that <t> is greater than it would be if the 
transformer were working on a non-inductive load; hence, 
the primary current corresponding to a given output will be 
greater with an inductive load. 

' The above diagrams represent the action of an ideal trans- 
former, and, as mentioned before, actual transformers 
approximate quite closely to the ideal. We will now notice 
what effect the resistance of the coils, core losses, and mag- 
netic leakage have on the action of an actual transformer. 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 11 



EFFECT OF RESISTANCE OF PRIMART ANB 
SECONDARY COILS 

16. Transformer coils always have an appreciable resist- 
ance ; hence, there v/ill be a loss in them proportional to the 
square of the current. One effect, therefore, of coil resist- 
ance is the heating under load and a lowering of the effi- 
ciency. A transformer with an excessive amount of coil 
resistance cannot have a high efficiency. The resistance 
also produces another effect that is more detrimental than 
the mere loss in efficiency, namely, bad regulation. When 
a transformer regulates badly, the secondary E. M. F., 
instead of remaining nearly constant, drops off as the load is 
applied and rises when it is removed. This is a particularly 
bad feature, especially when incandescent lights are being 
operated, because the changes in voltage not only affect the 
brilliancy of the lamps, but also shorten their life. The 
resistance of the primary prevents the induced back E. M. F. 
from being quite equal to the line E. M. F., because a cer- 
tain part of the impressed voltage is used in overcoming the 
resistance; this, in turn, will also cause the secondary 
induced E. M. F. to be slightly smaller than it would 
be if the primary had no resistance. The E. M. F. 
obtained at the terminals of the seco»dary will be further 
reduced by the drop due to the secondary resistance. It 
is thus seen that the general effect of the resistance is 
to cause a falling off in the secondary voltage when the 
transformer is loaded. The only way to prevent bad 
regulation from this source is to make the resistance of 
the coils as low as possible without making the design 
bad in other respects. 



EFFECT OF MAGNETIC LEAKAGE 

17« When a transformer has a large magnetic leakage, 
quite a number of the lines that pass through the primary 
will not thread the secondary ; consequently, the E. M. F. 

44—34 



Digitized by VjOOQ IC 



12 ' ALTERNATING-CURRENT APPARATUS §19 

induced in the secondary will not be as large as it should 
be. Take the case of a transformer constructed as shown 
in Fig. 1. When it is not loaded, there is no current in the 
secondary coil, and consequently there is nothing to oppose 
the primary coil setting up lines throuc^h the magnetic cir- 
cuit. Under these circumstances there would be very little 
leakage. When, however, a current flows in the secondary, 
a counter magnetic flux is set up that is opposed to the 
original- flux set up by the primary, as indicated by the 
arrows. There is then a tendency for poles to form at 
a\ a' and b\ b\ thus causing leakage lines b b and c c 
to be set up. Evidently the leakage will increase as the 
load on the secondary increases; hence, the tendency of 
magnetic leakage is to cause a falling off in the secondary 
voltage as the transformer is loaded. The effect of both 
resistance and magnetic leakage is, therefore, to produce 
bad regulation. 

18. Magnetic leakage can be avoided to a large extent 
by so placing the coils with reference to each other that 
all the lines passing through one must pass through the 
other. This might be done by winding the two coils 
together, but this plan would not work in practice, owing 
to the difficulty of maintaining proper insulation between 
them. The type shown in Fig. 1 would have a large 
amount of leakage and would not be used in practice. A 
much better arrangement would be to wind the coils one on 
top of the other, thus leaving very little space for lines to 
leak through between them, or to wind the primary and 
secondary in sections and interleave them. Transformers 
in which this is done will be illustrated later. 



EFFECT OF CORE LOSSES 



19. Since the magnetism in the core is constantly 
changing, there will be a hysteresis loss, just as there is a 



Digitized by VjOOQIC 



§ 19 ALTERNATING-CURRENT APPARATUS 13 

loss due to the varying magnetization of an armature core. 
The transformer will have to take power from the line to 
make up for this loss, thus lowering the efficiency. The 
amount of loss due to hysteresis depends on the quality and 
volume of iron in the core as well as on the maximum 
magnetic density at which it is worked. Since the mag- 
netic flux ^ is nearly constant for all loads, the magnetic 
density must also be constant, and the hysteresis loss must 
be about the same, no matter what load the secondary is 
carrying. Heat losses also occur in transformer cores, due 
to eddy currents set up in the iron. The iron in the core 
acts as a closed conductor, and the alternating field induces 
small E. M. F.'s, which give rise to currents in the core. In 
order to prevent the flow of eddy currents, the core is lam- 
inated or built up of sheets varying in thickness from 
.014 inch to about .025 inch, depending on the frequency, 
the thicker iron being used for transformers designed 
to work on low frequencies. The effect of both eddy- 
current and hysteresis losses is simply to increase the power 
that the primary takes from the line, and thus lower the 
efficiency. These losses do not affect the regulation to any 
appreciable extent, but if large they may lower the effi- 
ciency considerably. 

20, It has been shown abovfe that the core losses take 
place as long as the primary pressure is maintained, ^nd 
are about the same whether the transformer is doing 
any useful work or not. In many lighting plants, the 
line pressure is maintained all day, while the load may 
be on for only a few hours out of the twenty-four. It 
follows, therefore, that the PR, or copper, losses take 
place for a short time only, whereas the iron losses go 
on all day. In such cases it is important that the iron 
losses be small as compared with the copper losses, 
because, if this is not so, the transformer will have a low 
all-day efficiency; that is, it will give out a small amount 
of energy during the day compared with the amount it 
consumes. 



Digitized by VjOOQIC 



U ALTERNATING-CURRENT APPARATUS § 19 



COKSTRUCTIOK OF TRANSFORMERS 
-21. Transformers are made in a variety of forms, but 

they may, for conve- 
nience, be divided into 
two general classes: (1) 
Core transformers ; (2) 
shell transformers. 

In core transform- 
ers, the iron part forms 
a core on which the 
coils are wound, while 
in shell transformers 
the iron surrounds the 
coils. Figs. 5 and 6 
show the arrangement 
of the parts of a com- 
mon type of core trans- 
former. The core C, 
Fig. 5, is built up of 
thin iron strips into 
the rectangular form 
shown ; P, P\ 5, S' are 
the primary and second- 
ary coils, each wound in 
two parts. It will be 
noticed that the pri- 
mary is wound over 

the secondary, thus making the leakage path between 

the coils long and of 

small cross-section, ^_y^ Sec^mdary^ 

thereby reducing the 
magnetic leakage. 
Fig. 6 shows a section 
of the coils and core. 
One advantage of this 
type is that the core 
may be built up of strips of iron, no special stampings 





o 






p 














; i ■ 1 
i 

III 

II ' 
!! * 

1 1 
1 1 1 


II 








1 













Digitized by VjOOQ IC 



§ 19 ALTERNATING-C'URRENT APPARATUS 15 

being required. There is also an advantage in having 
the coils wound in two sections, in that it enables the 
transformer to be connected up for a variety of voltages. 



SXAMPI^S OF TRANSFORMERS 

23, Fig. 7 shows a small 2-kilowatt Westinghouse trans- 
former removed from its case. This transformer belongs to 
the so-called shell type, because the iron core is arranged 
so as to surround the coils. The iron stampings are shaped 



PlO. 7 



so that they can be easily slipped into place, and the joints 
overlap so that there is practically no break in the mag- 
netic circuit; the various shapes of stampings used for 



Digitized by VjOOQIC 



16 ALTERNATING-CURRENT APPARATUS § 19 

transformer cores tirill be considered in connection with 
transformer design. In this case, the primary is wound 

in two coils P^ P, Fig. 8, and 
the secondary coil 5 is placed 
between the two in order to 
reduce the magnetic leakage 
between primary and secondary. 
The terminals of the' primary are 
^'®- 8 connected to the posts /, /, /, /, 

Fig. 7, and by means of the pieces / the coils can be 




Fig. 9 



connected either in series or in parallel, so as to adapt 
the primary to either 1,050 or 2,100 volts. The secondary 



Digitized by VjOOQIC 



§ 19 ALTERNATING-CURRENT APPARATUS 17 

coil is also provided with a number of terminals, so that 
secondary voltages of 52J, 105, or 210 volts may be 
obtained. Fig. 9 shows a larger size of Westinghouse 
transformer. This is also of the shell type and has a 
capacity of 575 kilowatts. It is intended for use in 






Pio. 10 

connection with substations where a large transformer is 
required. The flat coils are spread apart where they pro- 
ject from the core, so that the heat developed in the wind- 
ings may be readily dissipated. Fig. 10 shows a section of 
a transformer of the same type as Fig. 7, but of larger size, 
mounted in its case and immersed in oil. 



Digitized by VjOOQIC 



18 ALTERNATING-CURRENT APPARATUS § 19 



23, The General Electric Company's type H trans- 
former is a good example of the core type. The general 

arrangement of the coils and core 
is shown in Figs. 5 and 6. Fig. 11 
shows the shape of the core and the 
arrangement of the coils in the 
larger sizes. The core A is built 
up of straight iron strips of two 
different widths. This leaves 
openings or ducts a at each 
corner for the oil, in which the 
transformer is immersed, to cir- 
culate through. An annular 
space d is also left between the 
coils r, ^, the object being to 
^'G. 11 secure thorough insulation and a 

free circulation of oil so as to keep all parts of the coil and 
core at approximately the same temperature and thereby 
avoid strains on the 
insulation due to un- 
equal expansion and 
contraction. The 
use of oil in trans- 
formers not only 
insures better insula- 
tion but it conducts 
the heat from the 
various parts to the 
outer case and thus 
makes the trans- 
former run cooler. 
Fig. 12 shows the 
style of cast-iron case 
used for holding 
type H transformers of moderate size which are usually 
mounted outdoors. The case is oil and water-tight and 
the primary leads a, b and the secondary leads c, d, e, /, 
are brought out through porcelain insulating bushings. 




Pio. tst 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS Id 

Substation transformers of large size must usually be 
provided with artificial means for cooling, which will be 
described later in connection with substations. 




PIO. 18 



24. Autotransformer. — Sometimes a transformer hav- 
ing but one winding, which serves both for the primary 
and secondary coils, is used for 
special purposes. A device of this 
kind is known as an antoti-ans- 
forraer. In Fig. 13, A represents 
a laminated iron core on which two 
coils having turns /, /' are wound. 
These two coils are connected in 
series so that they practically form 
one coil. The primary line wires 
are connected to the terminals ^, by 
and the secondary line wires to c 
and a. The pressure -£", depends on 
the number of turns in the coil /'. 
For example, if /' were one-third 
the total number of turns between b and ^, the voltage E, 
would be one-third the line voltage and the current in the 
secondary would be about three times that taken from the 
primary lines. 

The autotransformer is not suitable for use in connection 
with ordinary light and power distribution, because the sec- 
ondary is in direct electrical connection with the primary, 
and as the primary pressure is usually very much higher 
than the secondary, it would be dangerous to have the 
secondary system of wiring in connection with the primary. 
On this account autotransformers are only used for special 
purposes, where there is no great difference between the 
primary and secondary pressure, or where the use of the 
device is such that the electrical connection of the primary 
and secondary does not introduce an element of danger. 
Autotransformers, for a given capacity, are considerably 
cheaper to build than ordinary transformers, but their use 
is comparatively limited for the reason just given. 



Digitized by VjOOQIC 



20 



ALTERNATING-CURRENT APPARATUS § 19 



Lin* 



\M 



mm. 



8 



W 



PlO. 14 



Autotransformers are sometimes used on low-^pressure 
alternating-current switchboards to supply current to syn- 
chronizing lamps or other auxiliary devices. When used in 
this way, they are sometimes spoken of as s/iunt transformers, 

25. Series-Transformers. — Most transformers are run 
on constant-potential circuits, and their primary coils are 

connected across the 

^^ circuit. In some cases, 

however, series- trans- 

formers or current 

transformers are used 
for special purposes. 
These transformers 
have their primary con- 
nected in series with 
the circuit, as indicated 
in Fig. 14, where the 
coil P is the primary. 
The secondary is connected to whatever device A the cur- 
rent is to be supplied. As the current through the primary 
increases, the magnetization in the core will also increase, 
thus increasing the voltage in the secondary. If the resist- 
ance of A is fixed, the secondary current will increase in 
proportion to the secondary voltage, and hence in propor- 
tion to the current in the main circuit. One use of series- 
transformers has aleady been referred to in connection with 
the compounding of alternators. 

26. Series-Transformer for Ammeter. — Another very 
common use of the series-transformer is in connection with 
alternating-current ammeters. On account of the effects 
of self-induction, it is not practicable to use shunts with 
alternating-current ammeters, as is done with direct-current 
instruments. Shunts can be used with instruments that 
have practically no inductance, but with instruments of the 
plunger or dynamometer types the use of a shunt would 
interfere with the accuracy of the indications. In order to 



Digitized by VjOOQ IC 



§19 ALTERNATING-CURRENT APPARATUS 21 

reduce the current for these ammeters, a series-transformer 
is connected as shown in Fig. 14. The secondary can be 
wound so as to supply a comparatively small current to the 
ammeter A, and as this current is proportional to the main 
current, the dial of the instrument can be marked to indi- 
cate the current in the main circuit. Another, and in some 
cases a very great, advantage of the series-transformer is 
that it completely separates the instrument A from the 
high-pressure lines. 

Since the main current is usually quite large, a very 
small number of turns is sufficient on the primary P; in 
fact, in some cases, a single 
turn, or even a fraction of a 
turn, is enough. In some 
series-transformers used with 
ammeters, the secondary coil 
is wound on a laminated iron 
core built up of thin annular 
rings. This coil is simply 
slipped over the cable or other 
conductor carrying the main 
current. The current in the 
main conductor sets up an 
alternating flux in the lami- 
nated ring, and this sets up ^'^'- ^^ 
the current in the secondary. Fig. 15 shows the sec- 
ondary for a series-transformer of this type made by the 
Wagner Company. The iron ring with its winding is 
mounted on a fiber spool, and the main cable or other con- 
ductor carrying the current is passed through the hole in the 
center. The small flexible leads are connected to the 
ammeter. When series- transformers are used on very 
high-pressure circuits, they are generally mounted in cases 
in the same way as ordinary transformers, and immersed 
in oil. 

27. Potential Transformers. — It is not usual to con- 
nect voltmeters of the ordinary type directly across the 



Digitized by VjOOQIC 



PIO. 16 



n ALTERNATING-CURRENT APPARATUS §19 

line on alternating-current boards, because the pressure 

is so great that a voltmeter would require an exceedingly 
high resistance to permit its being so con- 
nected. Of course if the pressure were 
low, they could be connected in the ordi- 
nary way. In case the pressure is high, 
a small potential transformer is used 
to step-down the voltage. Fig. 16 shows 
a transformer of this kind. It is gener- 
ally mounted on the back of the switch- 
board; its primary coil is connected to 

the line and its secondary to the voltmeter, as shown in 

Fig. 17. It is bad practice to run switchboard lamps from 

the potential transformer, 

because, as a rule, the 

transformer does not have 

sufficient capacity for this 

purpose, and, besides, it is 

liable to interfere seriously 

with the accuracy of the 

voltmeter readings. The 

voltmeters are usually 

graduated to read the sec- 
ondary voltage, as this is 

what is generally required. In some cases, however they 

are graduated to indicate the primary voltage. 



Line 




FIO. 17 



Digitized by VjOOQIC 



§ 19 ALTERNATING-CURRENT APPARATUS 23 



ALTERNATING-CITRRENT MOTOES 

28. Motors designed for use in connection with alterna- 
ting currents may be divided into two classes: (1) Synchro- 
nous motors; (2) induction motors. 

Both kinds are in common use, and by far the larger part 
of all the motors operated in connection with alternating cur- 
rent belong to one of these classes. There are a few other 
motors that are used to some extent, but their number is 
insignificant compared with those of these two classes. 



SYNCHRONOUS MOTORS 

29. Synclironons motors are made to operate either 
on single-phase or polyphase systems, and are so-called 
because they always run in synchronism with, or at the 
same frequency as, the alternator driving them. In con- 
struction they are almost identical with the corresponding 
alternator, and always consist of the two essential parts, 
field and armature, either of which may revolve. The field 
of such motors must be excited from a separate continuous- 
current machine in the same way as an alternator. The 
fields of synchronous motors are, however, seldom if ever 
compound-wound, and hence are provided simply with col- 
lector rings, no rectifier being required; otherwise, the 
whole construction of the motor is about the same as that 
of the alternator. 

30. If a single-phase alternator be connected to another 
similar machine, the latter will not start up and run as a 
motor, because the current is rapidly reversing in its arma- 
ture, thus tending to make it turn first in one direction and 
then in the other. The consequence is that the armature 
does not get started from rest. If, however, the second 
machine be first run up to a speed such that the frequency 
of its alternations is the same as that of the alternator, and 



Digitized by VjOOQIC 



24 ALTERNATING-CURRENT APPARATUS § 19 

then connected in circuit, the impulses of current will tend 
to keep it rotating, and the machine will continue running 
as a motor. The motor must be run up to synchronism by 



^ \ \ 




means of some outside source of power, and the fact that 
single-phase synchronous motors will not start of their own 



Digitized by VjOOQ IC 



§19 ALTERNATING-CURRENT APPARATUS 25 

accord is a serious drawback to their use : in fact single-phase 
synchronous motors are now seldom installed. On the other 
hand, polyphase synchronous motors will start from rest 
and run up to synchronism when their armature windings 
are supplied with current, although in doing so they take a 
large current from the line, and if the motor is a large one, 
it is better to bring it up to speed by means of some outside 
source of power, such as an auxiliary motor. When current 
is supplied to the armature of a polyphase synchronous 
motor, magnetism is set up in the pole pieces by the arma- 
ture currents, and on account of hysteresis this magnetism 
lags behind the current. The consequence is that the mag- 
netism set up by the windings of one phase is reacted on by 
the current in the following phase, thus producing a torque^ 
on the armature and causing the motor to start up. If the 
pole pieces are not laminated, the volume of eddy currents 
set up in them may be considerable, and these will also aid 
in producing a turning moment on the armature; after the 
machine has come up to synchronism, its fields are excited by 
an exciter in the same way as an alternator. As stated 
above, this method of starting by simply connecting the 
armature to the line results in a large starting current, and is 
objectionable because of its disturbing effects on other 
parts of the system. For this reason, large motors are fre- 
quently started by means of a small induction motor that 
brings the large machine up to speed without taking a large 
line current. After the large motor has-been brought up 
to synchronism, the starting motor is disconnected by 
means of a clutch, and is then shut down. Fig. 18 shows a 
General Electric, 1,000-horsepower, two-phase, synchronous 
motor of the revolving-field type. It will be noted that the 
construction of the motor is the same as that of a revolving- 
field alternator. The large motor is brought up to speed 
by the small induction motor A, and after synchronism has 
been attained, the clutch B is thrown out and A is shut down. 

31. Synchronous motors behave differently in some 
respects from direct -current machines. If the field of a 



Digitized by VjOOQ IC 



26 ALTERNATING-CURRENT APPARATUS § 19 

direct-current motor be weakened, the motor will speed up. 
If the £eld strength of a synchronous motor be changed, 
the speed cannot change, because the motor must keep in 
step with the alternator. Such a motor adjusts itself to 
changes of load and field strength by the changing of the 
phase difference between the current and E. M. F. 
Imagine a synchronous motor which, we will suppose, runs 
perfectly free when not under load. If such a machine 
were run up to synchronism, and its field adjusted so that 
the counter E. M. F. of the motor were equal and oppo- 
site to that of the dynamo, no current would flow in the 
circuit when the two were connected. At any instant 
the E. M. F. causing current to flow is the difference 
between the instantaneous E. M. F. of the alternator and 
the counter E. M. F. of the motor. If the motor be 
loaded, its armature will lag a small fraction of a revolution 
behind^that of the alternator, and the motor E. M. F. will 
no longer be in opposition to that of the alternator; conse- 
quently, there will flow a current sufficiently large to 
enable the motor to carry its load. The greater the load 
applied, the larger will be the current that is thus allowed 
to flow. It must be borne in mind that this phase differ- 
ence is caused by a small relative lagging of one armature 
behind the other, not by a difl^erence in speed. For 
example, the change of phase from full load to no load 
might not be more than 25**, and this would mean an 
angular displacement on the machine of a little more than 
one-eighth of the pole pitch. If the machine be loaded too 
heavily, the slipping back of the motor armature will 
become sufficiently great to throw the motor out of 
synchronism, and it will come to a standstill. 

32. The action of a synchronous motor may be repre- 
sented as shown in Fig. 19. A represents the E. M. F. 
supplied to the motor. If the motor is running in synchro- 
nism, and if its voltage were almost equal and opposite to 
that of the alternator, its pressure would be represented 
by the dotted line B' lagging 180° behind A. Under such 



Digitized by VjOOQ IC 




§ 19 ALTERNATING-CURRENT APPARATUS 27 

circumstances very little current would flow, and the motor 
would exert but a very small torque. When the motor is 
loaded, the armature slips back a small 
fraction of a revolution, thus making the 
counter E. M. F. B of the motor lag by an 
angle a greater than 180°. The E. M. F. 
that will now be effective in forcing current 
through the circuit is the resultant of A and 
B^ and is represented by E. The current 
that will flow may be represented by the line 
/lagging behind the E. M. F. by the angle ^ 
that depends on the inductance and resist- 
ance of the motor circuit. It is easily seen 
that if a increases, the effective E. M. F. E 
also increases, and thus increases / to an 
extent sufficient to enable the armature to 
carry its load. The value of the E. M. F. B 
of the motor can be changed by changing 
the field excitation of the motor, while the 
speed is fixed by the speed of the alternator. 
By adjusting the field excitation of the motor, the current 
taken by it may be brought alpiost exactly into phase with 
the E. M. F. of the line, no matter what the load may be. In 
other words, a synchronous motor may be adjusted so as 
to run with a power factor of unity, and this is a consider- 
able advantage, especially where large motors are used. 
Further, if the field excitation is increased beyond that 
required to produce a power factor of unity, the current 
can, with an unloaded or lightly loaded motor, be made to 
lead the E. M. F. The motor under such conditions acts 
like a condenser of large capacity, and can, therefore, com- 
pensate for self-induction on the system, and improve the 
power factor of the system as a whole. Synchronous motors 
have been used to some extent in this way as compensators 
for self-induction, thus offsetting the bad effects of lagging 
currents and low-power factor. 

33« Synchronous motors are used mostly for work where 
the motor is not started and stopped frequently, and where 

44—35 



Pio. 19 



Digitized by VjOOQ IC 



28 ALTERNATING-CURRENT APPARATUS § 19 

it is not started under load. They are used mostly in the 
larger sizes. For ordinary work, involving frequent starting 
and stopping under load, induction motors are preferable. 

34. Speed and Direction of Rotation. — The speed at 

which a synchronous motor will run when connected to an 

2 n 
alternator of frequency « is J = — , where s is the speed 

in revolutions per second, and/ the number of poles on the 
motor. For example, if a ten-pole motor were run from a 
25-cycle alternator, the speed of the motor would be 

2 X 25 
s = — — — = 5 revolutions per second, or 300 revolu- 
tions per minute. It follows, then, that if the motor had 
the same number of poles as the alternator, it would run at 
exactly the same speed, and any variation in the speed of 
the alternator would be accompanied by a corresponding 
change in the speed of the motor. A synchronous motor 
will run in either direction, depending on the direction it is 
revolved when started up by its auxiliary motor. If, how- 
ever, it is started by simply allowing current to flow through 
the armature, its direction of rotation will depend on the way 
in which the armature terminals are connected to the line. 
Interchanging any two of the leads of a three-phase motor 
will reverse the direction of rotation, while interchanging 
the two wires of either phase of a two-phase motor will 
accomplish the same result. 



INDUCTION MOTORS 

36. In a great many cases it is necessary to have an 
alternating-current motor that will not only start up of its 
own accord, but one that will start with a strong torque. 
This is a necessity in all cases where the motor must start 
up under load. It is also necessary that the motor be such 
that it may be started and stopped frequently, and in general 
be used in the same way as a direct-current motor. These 
requirements are fulfilled by induction motors. 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 29 

36. Induction motors are usually made for operation on 
two-phase or three-phase circuits, although they are some- 
times operated on single-phase circuits, as explained later. 
They always consist of two essential parts, namely, the pri- 
mary^ or field, to which the line is connected, and the second- 
ary^ or armature, in which currents are induced by the action 
of the primary. Either of these parts may be the revolving 
member, but we will suppose that the field is stationary and 
the armature revolving, as this is the usual arrangement. 
In a synchronous motor or direct-current motor, the current 
is led into the armature from the line, and these currents, 
reacting on a fixed field provided by the stationary field 
magnet, produce the motion. In the induction motor, how- 
ever, two or more currents differing in phase are led into the 
primary, thus producing a magnetic field that is constantly 
changing and which induces currents in the coils of the arma- 
ture in the same way that currents are induced in the 
secondary coils of transformers. These induced currents 
react on the field and produce the motion of the armature. 
It is on account of this action that these machines are called 
induction motors. 

37. In order to understand the action of an induction 
motor, it will help matters to compare it briefly with the 
action of an ordinary direct-current motor. Suppose we 
have a direct-current armature surrounded by a four-pole 
field. If the field is excited and current sent into the 
armature through the brushes in the ordinary way, the 
armature will revolve, and the greater the load applied at 
the pulley, the more current will it take to drive the arma- 
ture. Suppose that instead of driving the armature in this 
way, we remove the brushes and press a copper ring over 
the commutator, so as to connect all the bars together. 
This will connect all the ends of the armature coils 
together, making them form a number of closed circuits. 
Also, suppose that we revolve the field around the armature 
instead of having it stand still, as is usually the case, and 
that the armature be held from turning. The lines of force 



Digitized by VjOOQ IC 



80 ALTERNATINO-CURRENT APPARATUS § 19 

from the field will cut across the armature conductors and 
set up E. M. F.'s in them. Since the coils are all short- 
circuited by the ring on the commutator, heavy currents 
are set up in them, and these currents reacting on the field 
produce a powerful dragging action on the armature. If, 
therefore, the armature is released, it will be dragged 
around after the field. If the armature revolved at exactly 
the same speed as the field, the conductors would move 
around just as fast as the lines of force, and hence they 
would not be cut by the lines of force, and no current or 
turning torque would be the result. It follows, then, that 
the armature must always revolve a little slower than the 
field, in order that any drag may be exerted. It should be 
noticed that in this arrangement no current is led into the 
armature from outside; it is induced in the armature by 
the revolving field. 

The field in this case is supposed to be excited by contin- 
uous current and revolved by mechanical means, but by 
using a two- or three-phase alternating current, we can 
make the magnetism sweep around the armature without 
actually revolving the field frame itself. In other words, 
we can set up magnetic poles that will be continually shift- 
ing around the armature without actually revolving the 
field structure. 

38. Revolving Meld. — The way in which two-phase 
currents can be made to set up a rotating magnetic field 
will be understood by referring to Fig. 20. This represents 
a simple form of field where the groups of conductors i, ^, 
etc. are laid against the inner surface of the laminated iron 
ring A, In an actual machine, these conductors would be 
laid in slots uniformly distributed, as shown in Fig. 21. 
This represents the stationary part of a Westinghouse 
two-phase induction motor, the coils a^ a being arranged 
around the inner circumference and making a winding very 
similar, so far as the arrangement of coils is concerned, to 
that used on direct-current armatures. These coils are 
connected up in the same way as those for a polyphase 



Digitized by VjOOQ IC 



§19 ALTERNATING-CURRENT APPARATUS 31 

alternator armature, and except for motors operating on 
high pressure, the windings are distributed uniformly 
instead of being bunched together into a few heavy coils. 




Cb> 







(d) 



(e) 



FIO. so 



In Fig. 20 we have taken a four-pole winding with only four 
groups of conductors for each phase. The conductors 
belonging to the two phases are marked 1 and ^, and the 



Digitized by VjOOQ IC 



32 ALTERNATING-CURRENT APPARATUS § 19 

different bands are separated by a small space so as to 
make the diagrams more easily understood. There are two 
flat coils for each phase, and these two coils are connected 




PIO. 81 



in series, so that the currents in those conductors that are, 
in this case, diametrically opposite flow in the' same direc- 
tion. Each coil consists of six turns. Fig. 20 (/) shows 
the ring with two of the coils in place, /, /' being the 
terminals of coil 2 2 and /", t'" those of coil 1 1. The other 
groups of conductors 2 2 and 1 1, shown unconnected so as 
not to confuse the figure, are connected up to form two 
coils similar to those drawn in; the coils marked ^ ;^ are 
connected in series as described above, and the terminals 



Digitized by VjOOQ IC 



§ 19 ALTERNATING^CURRENT APPARATUS 33 

connected to one phase of the two-phase circuit. The other 
two coils 1 1 are also connected in series and attached to 
the other phase. Let us now examine the nature of the 
magnetism set up when this field ring with its coils is pro- 
vided with two currents /, and /, differing in phase by 90°. 

39. In Fig. 20, the small arrows /, and /, represent the 
maximum values of the current in the coils and their pro- 
jection on the vertical dotted line gives the value of the 
current at any particular instant. In (a) the projection of 
/, is zero, while /, is at its maximum; hence, at this instant 
the current in conductors i^ ^ is at its maximum, and there 
is no current in conductors 1. From the way in which the 
coils 2 2 are connected current will flow up and down alter- 
nately in the groups of conductors marked 2, those in 
which the current flows up through the plane of the paper 
being marked with a dot, and those with a down-flowing 
current filled in black. Conductors 1 1 are left blank 
because at this instant no current flows in them. By 
remembering the rule governing the direction of current in 
a wire and the direction of the lines of force set up around 
it, it can easily be seen that the magnetism set up around 
the bands of conductors carrying the current will be as 
shown by the dotted lines. Where these lines leave the face 
of the ring, as indicated by the arrowheads, a north pole is 
formed, and where they enter the ring a south pole is 
formed, the direction of the lines and the current being 
related in the same way as the direction of turning and 
movement (downwards or upwards) of a right-handed 
screw. It is easily seen from {a) that four equally spaced 
poles are formed around the ring, the centers of these poles 
being indicated by the arrowheads in the center of the 
figure and also by the letters N S, 

In {h) the condition of affairs, \ cycle later, is indicated. 
.Here the instantaneous values of /, and /, are equal and in 
the same direction, as indicated by their projection on the 
vertical line. Also, the current in conductors 2 is still in 
the same direction as in («). The magnetism set up around 



Digitized by VjOOQ IC 



84 ALTERNATING-CURRENT APPARATUS § 19 

the conductors will now be as indicated, and the four poles 
have shifted around -^ revolution. At the instant shown 
in {c) /, has become zero, and /, has reached its maxi- 
mum value. The poles have shifted another ^revolution 
in the direction indicated by the arrow. In {d) we have the 
condition of affairs ^ cycle later, when /, and /, are equal 
but opposite in direction; /, has the same direction as in 
{d)y but /, is in the opposite direction. The four poles, 
therefore, occupy the position shown, thus shifting around 
^ revolution. One-eighth of a cycle later, /, is zero and 
/, at its maximum in the opposite direction, as shown 
in {e); this figure is the same as (^), except that the 
currents in conductors 2 2 are reversed and the poles shift 
forwards to the position indicated. During the time repre- 
sented by these figures, the current has passed through 
\ cycle, the poles have made \ revolution around the 
ring, and as the currents in the two phases continue to 
change, the magnetism sweeps around, although both the 
coils and core remain fixed. In the case shown, the field 
makes ^ revolution for every complete cycle of the current. 
If a six-pole winding were used, the field would make but 
J revolution for each complete cycle, or in general 

where s = speed of revolving field in revolutions per second; 
/ = number of poles; 
n = frequency in cycles per second. 

40. Armatures for Induction Motors. — Referring 
again to the direct-current machine, we can replace the 
mechanically revolved field magnet by a revolving field pro- 
duced by means of polyphase currents, and if the armature 
with the short-circuited commutator be placed within such a 
field it will be dragged around as previously explained. It 
would, of course, be a needless expanse to provide a short- 
circuited commutator, so that this type of armature would 
never be used. Fig. 22> shows a very common type of 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 35 

induction-motor armature intended for use in connection 
with the field shown in Fig. 21, and its construction is 
exceedingly simple. The laminated core is provided with a 
number of slots around its periphery, in each of which an 



Pig. X8 

insulated copper bar b is placed. At each end of the arma- 
ture is a heavy copper ring r, to which the projecting ends 
of the bars are bolted, thus connecting all the bars together 
and making them form a number of closed circuits. This 
arrangement is called a squirrel -cage armature. When 
an armature of this kind is placed in a revolving field, 
the rotating magnetism sets up E. M. F.'s in the arma- 
ture conductors, thus causing currents to flow through the 
closed circuits and exerting a torque on the armature. The 
iron core of the armature completes the magnetic circuit of 
the field, so that the magnetism instead of passing through 
the air, as shown in Fig. 20, passes through the iron of the 
armature core, thus making the magnetic circuit consist 
wholly of iron with the exception of the small air gap 
between armature iind field. 



Digitized by VjOOQ IC 



36 ALTERNATING-CURRENT APPARATUS §19 

41, Slip. — If the armature were held from turning in 
a revolving field, the coils on the armature would act like 
the secondary of an ordinary transformer, and heavy cur- 
rents would be set up in them. However, as the armature 
comes up to speed, the relative motion between the revolving 
field and armature becomes less, and the induced E. M. F/s 
and currents become smaller, because the secondary turns do 
not cut as many lines of force as before. If the armature 
were running exactly in synchronism with the field, there 
would be no cutting of lines whatever, no currents would be 
induced, and the motor would exert no torque. Therefore, 
in order to have any induced currents, there must be a dif- 
ference in speed between the armature and the revolving 
field, and the greater the current and consequent torque or 
effort, the greater must be this difference. When the load 
is very light, the motor runs very nearly in synchronism, 
but the speed drops off as the load is increased. This differ- 
ence between the speed of the armature and that of the field 
for any given load is called the slip. The slip in well- 
designed motors does not need to be very great, because 
the armatures are made of such low resistance that a small 
secondary E. M. F. causes the necessary current to flow. 
In well-designed machines it varies from 2 to 5 per cent, 
of the synchronous speed, depending on the size. A 
20-horsepower motor at full load might drop about 5 per 
cent, in speed, while a 75-horsepower motor might fall off 
about 2 J per cent. For example, if an eight-pole motor 
were supplied with current at a frequency of 60, its field 
would revolve ^ = 15 revolutions per second, or 900 revo- 
lutions per minute, and its no-load speed would be very 
nearly 900. At full load the slip might be 5 per cent., so 
that the speed would then be 855 revolutions per minute. 
It is thus seen that as far as speed regulation goes, induc- 
tion motors are fully equal to direct-current machines. If 
5' represents the speed of the armature and 5 the speed of 
the revolving field in revolutions per second then 

slip = 5-5', (6) 



Digitized by VjOOQIC 



§ 19 ALTERNATING-CURRENT APPARATUS 37 

or exjiressed as a percentage of the speed that the armature 
would run at if it were in synchronism with the field, 

slip (per cent. ) = ^^ ^r-^ (6) 

Since the armature revolves very nearly as fast as the 
rotating magnetic field, the currents in the armature are of 
correspondingly low frequency. The greater the slip, the 
greater is the frequency of the currents in the secondary. 
If the secondary were held from turning, the slip would be 
100 per cent., and the frequency of the secondary currents 
would be the same as that of the primary. If the slip were, 
say, 5 per cent., and the primary frequency 60 cycles per 
second, the frequency of the secondary currents would be 
3 cycles per second. On account of the fact that induction 
motors do not run in synchronism with the source of supply, 
they are often called asynchronous motors, to' distinguish 
them from the synchronous type. 

. 43. In speaking of induction motors, the stationary part 
is often referred to as the staler, and the rotating part as 
the rotor. In order to avoid the use of slip rings, the field, 
or primary, into which the currents are led from the line is 
usually the stator, while the armature, or secondary, in 
which the currents are induced is the rotor. The terms 
primary and secondary are used in reference to the field 
and armature, because of the similarity between the action 
of the induction motor and the ordinary transformer. 

43. Relation Between Torque and Slip. — If the 

armature ran in synchronism with the field, the torque 
would be zero, as explained above, because there would then 
be no current in the armature. As the slip increases, the 
torque also increases, and it would increase in direct propor- 
tion were it not for the demagnetizing effects of the currents 
in the armature. The currents in the armature oppose the 
magnetizing action of those in the field, and as the armature 
current increases, this opposing action causes part of the 
magnetism to leak across the space between the primary and 



Digitized by VjOOQ IC 



38 ALTERNATING-CURRENT APPARATUS § 19 

secondary in much the same way as magnetic leakage occurs 
in an ordinary transformer. The result is that as the slip 
and the armature currents increase, the demagnetizing 
action and magnetic leakage also increase, so that while the 
armature current is large, the flux through the armature 
becomes smaller and the torque may actually become less 
instead of greater. For every motor there is a certain slip 
at which the maximum torque is exerted, and beyond which 
the reaction of the armature becomes such that the torque 
decreases with increase of slip. This point depends on the 
construction of the motor. The torque at starting, i. e., 
with slip 100 per cent., depends largely on the resistance of 
the armature. If the armature has a very low resistance, 
the current in it at starting will be very large, the armature 
reaction will be large, and the resulting torque small. If, 
on the other hand, the armature resistance is high, the 
current will be limited in amount and the starting torque 
will be high, because the current that flows in the armature 
will have a strong field to react on. An induction motor 
with a high armature resistance exerts its maximum torque 
at a low speed, while one with a low armature resistance 
exerts its maximum torque at a high speed. It is thus seen 
that by varying the construction of a motor of given size 
and output, its performance as regards torque, speed, etc., 
can be changed considerably. 

The curve in Fig. 23 shows the relation between the torque 
and speed of a typical induction motor. The point o repre- 
sents a slip of 100 per cent., i. e., the motor is at a standstill, 
and point a represents a slip of zero, i. e., synchronous speed. 
When the motor is running at synchronism, the torque is 
zero, and as the speed falls off from synchronism the torque 
rapidly increases until the maximum value ^ ^ is reached. 
With further increase in the slip, the torque falls off until 
at a standstill it has fallen to the value o rf, which would 
represent the torque at starting. 

44. Induction Generator. — If the motor is to be driven 
above synchronism, a torque must be applied in the negative 



Digitized by VjOOQIC 



§ 19 ALTERNATING-CURRENT APPARATUS 39 

direction, or, in other words, power must be supplied to the 
motor from an outside source, as indicated by the curve a ef. 
If driven above synchronism in this way, the induction 




PIO. 88 



motor would give back current to the line and would there- 
fore be an Induction generator. In order that an induc- 
tion generator may furnish current, it must be connected 
to an alternator or live circuit that is capable of exciting its 
field. Induction generators have not been used to any extent 
in practice, though in a few cases where induction motors 
have been used for operating electric cars, they have been 
made to act as generators to return current to the line when 
the cars descend grades, thus acting as an electric brake and 
at the same time returning energy to the system. If the 
motor were driven backwards, as represented by d k, then 
since the torque is in the same direction as before, the motor 
would simply act as a brake. The portion of the torque 
curve that represents the actual conditions under which the 
motor is run is that lying between ^and c. If the motor is 
loaded beyond the point represented by r, the torque will 
diminish and the motor will stop. 

46. Reversing Direction of Rotation. — In order to 
reverse the direction of rotation of an induction motor, it is 
necessary to reverse the rotation of the revolving magnet- 
ism set up by the field windings. In a two-phase motor 



Digitized by VjOOQ IC 



40 ALTERNATING-CURRENT APPARATUS § 19 

this can be done by reversing the current in either of the 
phases, i. e., by interchanging the connections of one of 
the phases with its terminals on the motor. A tliree-phase 
motor can be reversed by interchanging the connections of 
any two of the line wires with the motor terminals. 



46. Speed Kegrulation of Induction Motors. — As 

already stated, the induction motor tends to run nearly in 
synchronism with the alternator that supplies it with cur- 
rent. Its speed can never quite reach synchronism, because 
it always takes some power to make up for the friction 
losses, etc., even if the motor is unloaded. It is also 
evident that the speed cannot rise above that of synchro- 
nism, and that with the exception of the slight variations 
in speed, due to the changes in load and corresponding 
change in the slip, the speed of the motor remains practi- 
cally constant as long as the speed of the alternator and 
voltage on the line remain constant. Generally speaking, the 
induction motor is not as well adapted for variable speed as 

the direct-current motor, 
although its speed can be 
varied through a consider- 
able range. Speed regula- 
tion of induction motors is 
usually accomplished by pro- 
viding the armature with a 
regular three-phase Y-con- 
nected winding and connect- 
ing a variable resistance in 
series with each branch. 

Fig. 24 shows the method 
referred to. The bar winding 
on the armature is arranged 
in three phases tf, ^, r, Y-con- 
nected, instead of the simple 
squirrel-cage arrangement. 
The three terminals of the winding are connected to collector 
rings mounted on the shaft, and from them the armature 




FIG. 94 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 41 

current flows through the resistances r„ r„ r,. When the 
resistance arm d is in the position shown, all the resistance is 
in circuit, but when the arm is moved around to the dotted 
position, as indicated by the arrow, the three ends of the 
winding are connected together directly through rf, thus 
short-circuiting them. When there is a high resistance 
in the armature circuit, the slip must be large in order to 
make enough current flow to provide the requisite torque; 
consequently, when all the resistance is in, the speed of the 
motor will be low. Where motors are used for operating 
cranes or other hoisting machinery, it is necessary to have 
a variable speed, and this is usually accomplished in the 
manner just described. 



METHODS OF STARTING INDUCTION MOTORS 

47, An induction motor may be started by connecting 
its field directly to the line, but this allows a large rush of 
current, which, disturbs other parts of the system and is, 
therefore, objectionable. This method of starting is not 
practicable with any but small motors. In order to obtain 
a smooth start and thus avoid a rush of current, either of 
two methods may be adopted. The voltage applied to the 
primary may be reduced either by inserting a resistance or 
by the use of an autotransformer. Or a resistance may be 
inserted in the secondary at starting, and cut out when the 
motor comes up to speed. 

48, Starting^ Compensator, or Autotransformer. — 

Where a motor is provided with a squirrel-cage winding, it 
is generally started by cutting down the voltage applied to 
the primary. This is usually done by means of an auto- 
transformer inserted between the line and the motor field, 
the transformer being provided with a double-throw switch, 
so that it can be cut out when the motor has come up to 
speed. 

Fig. 25 shows a Stanley starting: compensator, auto- 
transformer, or autostarter, as it is variously called* This 



Digitized by VjOOQ IC 



42 ALTERNATING-CURRENT APPARATUS § 19 



starter is intended for a three-phase motor, and is equipped 
with three autotransformers, one for each phase. The 
coils and cores of the transformers are contained in the 
box M^ for which E is the cover. The switch contacts are 
controlled by the lever »S' and are arranged so that the 
circuits are always broken under oil contained in the recep- 
tacle D (which is here shown 
removed from its normal posi- 
tion). The motor is started 
by throwing the switch 5 from 
the off-position to the start- 
ing position; these points are 
plainly marked on the side of 
the box M, After the motor 
has come up to speed, the 
switch is thrown over to the 
running position, thus cutting 
the autotransformers out of 
circuit. Fig. 26 shows the 
connections. Wires a^ b, f, 
Fig. 25, connect to the three- 
phase supply mains, and A^ 
B, C are the coils of the auto- 
transformers, each of which is 
provided with a number of 
taps 7, 2, 3, ^ Fig. 26. When 
the switch is thrown to the 
left, coils A, B, C are con- 
nected in circuit with the sup- 
ply mains. The motor wind- 
ings are, however, connected 
only across that portion of each coil that lies between the 
points 1 and 5\ consequently, the voltage applied to the 
motor is decreased and the current is correspondingly 
increased. The voltage applied to tl>e motor at starting 
can be adjusted by using the taps 1, 2, 3, Jf, so that the start- 
ing current can be varied to suit the conditions under which 
the motor is used. A simple arrangement is provided so 




Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 43 

that leads /, /', /" can be connected to points i, 2, 3, or ^, 
as desired. If connection is made at point 4, the maxi- 
mum starting effort is obtained with a correspondingly 
large current taken from the line. The General Electric 
Company also uses a similar starting device in connection 



FIO.28 

with their two- and three-phase induction motors. The 
three-phase starter has three coils, and its principle of 
operation is the same as that just described. When 
starting motors with such devices, time should be allowed 
after the switch is placed in the starting position for 
the motor to come up to nearly full speed before turn- 
ing the switch to the running position, otherwise the 
fuses will be blown. Fig. 27 (a) shows the connec- 
tions of the General Electric three-phase starting com- 
pensator. 

In Fig. 27 {a) a main switch is shown between the start- 
ing compensator and the line. This is not always installed, 
because the compensator itself can be made to answer the 
purpose of a switch ; however, it is better to have a main 
switch or circuit breaker, so that the compensator can 

44—36 



Digitized by VjOOQ IC 



7b Lm9 o/* 



Running 
Position 




Pig. 97 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 45 

be completely cut off from the line if occasion demands. 
An induction motor if overloaded excessively will stop and 
will soon overheat unless the current is cut off. In order 
to protect these motors, either fuses or circuit breakers may 
be used, though sometimes they are installed without any 
protective device. The fact that the motor stops if loaded 
excessively is often depended on to serve as an indication 
of overload, but this method cannot be recommended, and 
it is better to have some form of protective device. 

The trouble with fuses and circuit breakers, if installed in 
the usual manner, is that they may open the circuit when 
the motor is being started, because there is always a large 
current for a short interval, even if everything is all right. 
The protective devices thus act when they are not wanted 
to, and for this reason they are often omitted. In order 
to overcome this objection, the fuses, see Fig. 26, may be 
connected so that they will not be in circuit during the 
starting interval, but will be cut in during the time the 
motor is in operation. Fig. 27 (6) shows a three-phase 
motor with its starting compensator protected by a triple- 
pole Cutter circuit breaker (a detailed description of this 
circuit breaker will be found in a later section). The circuit 
breaker opens all three lines, but the tripping coils are only 
in circuit when the compensator switch is in the running 
position. An overload during the regular operation of the 
motor will therefore cause the breaker to open the circuit, 
whereas the large starting current will not. 

In some cases a low starting voltage can be obtained 
without the use of a compensator. For example, the step- 
down transformers that supply the motor can have taps 
brought out from the middle point of their windings and 
connected to a double-throw switch, so that in one position 
of the switch the motor gets only half the normal voltage, 
while for the running position of the switch it is connected 
across the secondaries in the usual manner and gets the full 
voltage. Another scheme that has been used with two- 
phase alternators having the two armature windings inter- 
connected, as, for example, in Westinghouse two-phase 



Digitized by VjOOQ IC 



46 ALTERNATING-CURRENT APPARATUS § 19 

machines, is to provide a double-throw switch, so that at 
starting the motor is connected across the *'side phases." 
Instead of receiving the full-line voltage -£", it then hjis a 

E 
voltage — = applied to it, and after it has come up to speed, 

the switch is thrown over and the motor receives the full-line 
voltage E of each phase. This method is used considerably 
in shops where induction motors are supplied with current 
directly from the alternator. 

49. Starting With Resistance in Armature. — The 

compensator method of starting has the advantage of being 
simple and allowing the use of a squirrel-cage armature, 
which is easy to construct and not liable to get out of order. 
Also, the starter can be placed in any convenient location. 



Pig. 28 

chus permitting the motor to be controlled from a distant 
point. On the other hand, the torque of an induction 
motor decreases very rapidly when the applied voltage is 
decreased; it varies as the square of the applied voltage, 
and if a strong starting torque with a moderate current 
from the line is desired, it is better to use a resistance in 
the armature when starting, and apply the full voltage to 



Digitized by VjOOQ IC 



1 19 ALTERNATING-CURRENT APPARATUS 47 

the field. Since the starting resistance is only in use for a 
short time, it may be made of sufficiently small bulk to be 
mounted on the supporting spider of the armature, thus 
doing away with the collector rings shown in Fig. 24. Where 
the resistance is used for speed -regulating purposes, con- 
siderable heat is generated in it all the time the motor is 
running; this heat would be objectionable in the machine, 
and, besides, a resistance for continuous use would usually be 
too bulky to mount within the armature. 

Fig. 28 shows an armature for a General Electric induc- 
tion rnotor in which the resistance is mounted on the spider. 
The resistance is in three parts a^ b^ Cy one part for each 




FlO. 99 

phase, and is made in the form of cast grids. It is cut out 
of the armature circuit by means of a sliding switch 
operated by a loose knob attached to a spindle running 
through the center of the shaft. The sliding contact that 
cuts out the resistance is attached to the end of this 



Digitized by VjOOQ IC 



48 ALTERNATING-CURRENT APPARATUS § Id 

spindle, and by pushing in the knob when the motor attains 
its speed, the resistance is cut out and the three phases of 
the armature winding directly short-circuited. By adopting 
this construction, the use of collector rings is avoided, but 
the construction of the armature as a whole is considerably 
more complicated than that of the squirrel-cage type. 

Fig. 29 shows a General Electric three-phase motor 
equipped with the style of armature shown in Fig. 28. To 
start a motor of this kind, all that is necessary is to see 
that the knob is out as far as it will go, so that all the 
resistance is in circuit; then throw in the main switch. 




PIO. 80 

The motor will start up with a good torque accompanied by 
a moderate line current, and will come up to speed in about 
15 seconds, during which time the knob is pushed in, thus 
cutting out the resistance. 

Fig. 30 shows one of the earlier types of General Electric 
motor in which the starting resistance is cut out by the 
sliding collar c operated by the lever h. 



Digitized by VjOOQ IC 



§19 ALTERNATINO-CURRENT APPARATUS 4d 



TIEUD CONNECTIONS 

60« As already pointed out, the Held connections of a 

multiphase induction motor are similar to those of a multi- 
phase alternator armature, and what has been said regard- 
ing alternator armatures applies to induction motor fields. 
The field windings of a three-phase motor may be connected 
either Y or A, depending on which is best adapted to the 
voltage and current with which the motor is to operate. 
If, for example, the field were to be connected to high- 
potential mains, it would probably have its coils con- 



rbj 

PlO. 81 

nected Y. Fig. 31 shows a simple arrangement of coils 
suitable for a three-phase motor field; the connections 
between the coils are not shown, as the figure is intended sim- 
ply to illustrate the grouping of the coils. The stampings F 
are provided with 24 slots, in which are placed 12 coils, 
a, b^ r, etc. The coils belonging to the three different 
phases are shaded differently, in order to distinguish them 
from one another. The four coils of each group may be 
connected in series or parallel, and the three groups then 



Digitized by VjOOQ IC 



50 ALTERNATING-CURRENT APPARATUS § 19 

connected together either Y or A. Fig. 31 (d) shows how 
the coils are arranged so that the ends will clear each other. 
This field winding is designed for an eight-pole revolving 
field, and the student should compare it with the corre- 
sponding alternator winding. 

61« The field winding of induction motors usually 
consists of several groups of conductors per pole per phase, 
instead of a single group, as shown in Fig. 31 (a), and the 
winding becomes uniforml)' distributed, as shown in Fig. 21. 
Sometimes, however, the field is wound to take current at 
high pressure, and in such cases it is necessary to design 
the winding so that it will be as free as possible from cross- 
ings, and consist of a comparatively small number of coils 
that can be thoroughly insulated. For motors of this type, a 
winding similar to that shown in Fig. 31 would be suitable. 

62« Power Factor of Induction Motors. — Induction 
motors always give rise to lagging line currents; that is, 
the actual watts taken from the line are not equal to the 
volts X amperes, but this product multiplied by the 
pow^er factor of the motor. The higher the self-induc- 
tion of the motor and the higher the frequency, the lower 



r 



^""1^ 



IHJ 



(a) 



"YTTT^T^ 



1 



r 



v^ 



PIO. 88 



^e) 



1 



will be the power factor, other things being equal. Every 
effort is made to design induction motors so that the power 
factor will be high. The use of open slots, as at (a) and (6), 
Fig. 32, tends to keep down self-induction, because an air 
gap ^ ^ is introduced into the path of the magnetic lines that 
the coil tends to set up around itself. If closed slots are 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 51 

used, the inductance is greater, because the coils can set up 
a flux through a complete iron path. The power factor, 
cos <t>y of a good motor of medium size operating on 
60 cycles should be from .85 to .87 at full load. Motors 
of large size may have power factors of .9 and over. 

53, Induction motors are always constructed with a 
multipolar field, so as to keep down the speed of rotation. 
The number of poles employed increases with the output, 
and the speed is correspondingly decreased. The following 
table gives the relation between poles, output, and speed for 
some of the standard sizes of induction motors (60 cycle). 

TABIiE I 
INDUCTION MOTORS 



Poles 


Horsepower 


Speed 


4 


I 


1,800 


6 


5 


1,200 


6 


lO 


1,200 


8 


ID 


900 


8 


20 


900 


lo 


50 


720 


12 


75 


600 



64, Cbaracteristlc Curves of Induction Motor. — The 

general performance of an induction motor is best illustrated 
by means of curves showing the relation between output and 
speed, power factor, efficiency, slip, etc. Fig. 33 shows 
curves for a 30-horsepower 60-cycle 110- volt motor, given by 
Steinmetz. As the load is increased, the speed falls off and 
at full load, 30 horsepower, it has dropped about 3 per cent. 
The power factor increases with the load, and at full load it 
is about 83 per cent. As the load is still further increased, 
the power factor reaches a maximum of about 85 per cent. 
The apparent efficiency of the motor is the ratio of the watts 
output to the apparent watts input, and its value at full load 
is about 71 per cent. The actual efficiency of the motor is, 
however, considerably higher than this because the actual 



Digitized by VjOOQ IC 



82 ALTERNATING-CURRENT APPARATUS § 1ft 

input is less than the apparent, being equal to the apparent 
watts multiplied by the power factor. The actual efficiency 
at full load is about 86 per cent. The maximum load that 
the motor will deliver without stopping is a little over 
60 horsepower. This point is indicated by the bending of 



























N 




3 




























\ 






•1 












Brmmn 




— 









4#9 




s S 




.^^ 


^^ 




mrr 


ICIMMt 


r 




— = 


==c 










_^^.* — 





■—" 










/ 


^ 


A 


y^^ 














/ 


'^' 








/ 


/ 


A 


-'^^ 












y 


^< 


N 






/ 


/ 


'A 














y 


iT 




\ 
\ 
1 

/ 


^»A 




/ 1 


V 












,^ 


^ 








/ 
/ 

• 
• 








/ 
/ 
/ 
/ 








5SJJ 


JJJ^ 


















\ ' 


/ 




^ 




















l## 




V 


-^ 


























# 


!' 


f .1 


r 11 


F . .1 


r » 


or 
« 1 


trUTt 


m.r.- 


} i 


\ § 


> « 


§ i 


t fl 


• 



Pig. 88 



the curves. Of course, in actual work the motor would not 
be loaded to anything like this amount on account of the 
excessive heating that would result. 



SINGLE-PHASE rNDUCTION MOTORS 

55. If a motor constructed on the same lines as an 
induction motor, but provided with only a single winding on 
the field instead of two or more sets of windings differing in 
phase, were connected to single-phase mains, it would not 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 63 

start up of its own accord. If it were given a start by pull' 
ing on the belt, it would gradually come up to speed in 
whichever direction it was started, provided no load was 
applied. The motor would exert very little torque, but it 
would gradually increase in speed until it attained a speed 
nearly in synchronism with the generator. After the motor 
has been started, the load may be applied and the motor will 
behave in the same way as one operated on a polyphase sys- 
tem. A two-phase or three-phase motor will run on one 
phase after it has been started, though, of course, it will 
take more current in the single-phase than when all the 
phases were in use. 

When current is first applied to such a motor, it acts as 
an ordinary transformer, currents being set up in the sec- 
ondary. The magnetic field is set up by the joint magnet- 
izing action of the primary and secondary currents, as in 
the case of a transformer, and when the armature is given a 
start and made to revolve, these induced secondary currents 
are carried around in space with regard to the primary, so 
that a similar effect is produced as if the primary were pro- 
vided with windings displaced in phase. As the armature 
comes more nearly into synchronism, the combined action of 
the primary and secondary currents is to produce a rotating 
field very similar to the true rotating field set up by two or 
more windings supplied with polyphase current. The objec- 
tions to a motor of this kind are, of course, that it will not 
start up of its own accord and that it will not start under 
load ; it is only after the motor has attained its speed that it 
will carry a load and behave in general like a two- or three- 
phase induction motor. 

66. In order to make a single-phase motor start from 
rest and give a good starting torque, it is necessary to pro- 
vide a rotating field at starting. This can be done by using 
regular two-phase or three-phase windings and supplying 
these windings with displaced E. M. F.'s obtained hy splitting 
the phase, as it is called. Many different methods for phase 
splitting are in use. 



Digitized by VjOOQIC 



64 ALTERNATING-CURRENT APPARATUS § Id 

In Fig. 34 the motor is provided with a two-phase wind- 
ing, and in series with winding yi, a non-inductive resistance 



Line 




PlO. 84 



R is connected. An inductance L is connected in series with 
B^ and the two windings are connected in parallel across the 



ii-AAAAArPlM^« 




FIO. 35 

lines. It is evident that the current in B will lag behind 
that in A, and if the resistance and inductance are correctly 



Digitized by VjOOQIC 



g 19 ALTERNATING-CURRENT APPARATUS 55 

proportioned the currents can be made to differ enough in 
phase to produce an imperfect form of rotating field suffi- 
cient to start the motor. The windings are frequently so 
designed that the necessary phase displacement is caused 
by the windings themselves, and outside resistance and 
inductance rendered unnecessary. In some cases one 
of the windings is a main, or working, winding, and the 
other is used only at starting, being cut out by open- 
circuiting it by means of a switch after the motor has 
attained its speed. 

Fig. 35 shows another scheme for starting a motor on 
single-phase mains. Two of the terminals are connected to 
the mains, and the third terminal is connected to the point b 
between the resistance R and inductance L, The E. M. F. 
between a and b differs in phase from that between b and r, 
so that the different windings of the motor are supplied 
with displaced E. M. F.*s suitable for starting. A switch 
can easily be arranged to disconnect R and L after the 
motor has come up to speed, thus running it on the two 
outside lines only. 

Fig. 36 shows a 
starting arrange- 
ment similar to 
Fig. 35, except that 
a condenser C is 
used instead of the 
ind uct ance L, 
Sometimes where 
this combination is 
used, R and C are 
not cut out after the 
motor has attained 
speed, because the 
condenser C coun- 
teracts the self- 
induction of the 
motor and thus raises its power factor to such an extent 
that the small amount of loss in R is more than made up. 




PIO. 86 



Digitized by VjOOQIC 



56 ALTERNATING-CURRENT APPARATUS § 19 

57, It is thus seen that so far as construction is 
concerned, the single-phase induction motor is almost 
identical with the polyphase motor, the principal differ- 
ence being in the method of starting and the way in 
which the revolving field is set up after the motor is in 
operation. Fig. 37 shows an interesting starting arrange- 
ment used very largely for small single-phase fan motors. 
Each pole piece A is provided with a magnetizing coil D 
in the usual manner; a slot c is made in the pole piece 
in which is placed a rectangular copper stamping B. In 
some cases this copper stamping is replaced by a num- 
ber of turns of wire, the two ends of the coil being 
joined together so as to make a closed circuit. The 




PlO. 87 

field coil B sets up a magnetic flux a, and the flux i 
through the portion of the pole face covered by B is 
due to the combined effect of D and B. The flux pass- 
ing through the copper stamping or shading coil, as it 
is often called, sets up induced currents that are out 
of phase with the flux; the effect of these induced cur- 
rents is similar to that of a second set of coils with cur- 
rents in them differing in phase from the current in 
the main field coils. In other words, the flux in one 
part of the pole face differs in phase from that in the 
other, thus producing the effect of a shifting field to a 
sufficient extent to bring the motor up to speed. The 
armature in these motors is a simple squirrel cage with 
round copper bars. 



Digitized by VjOOQIC 



§ 19 ALTERNATING-CURRENT APPARATUS 67 

SEBIBS-MOTOB ON ALTERNATING CITRRENT 

68. If a motor constructed in every way like a series- 
wound direct-current machine be provided with laminated 
fields and supplied with current from alternating-current 
mains, it will start with a good torque and run up to speed 
under load, thus making a single-phase alternating-current 
motor. Since the field is in series with the armature, it 
follows that the current in each reverses at the same 
instant. It has already been shown in connection with 
direct-current motors that if the currents in both field and 
armature are reversed, the direction of rotation remains 
unchanged. Series motors, with laminated field, have here- 
tofore been used very little on alternating current circuits 
because of the difficulty of making motors that would run 
without bad sparking at the commutator. Recently, how- 
ever, much attention has been paid to the development of 
this type of motor with a view to its use for operating electric 
railways, and the design has been perfected so that motors 
of suitable size can now be built that will operate as well as 
series direct current motors of corresponding size. The 
series alternating-current motor has the same characteristics 
as the corresponding direct-current machine ; i. e. the 
speed increases with decrease in load, and at the same time 
the torque decreases. Large torque at starting can be ob- 
tained and the motor is thus particularly suited to railway 
work. The frequency on which these motors are operated 
must be low (25 Cycles or under), and as a frequency of 25 
has become standard in railway work, most of them have 
been designed for that frequency. One advantage of the 
motor is that it will operate on either direct or alternating 
current. On account of its variable speed it is not well 
adapted for most stationary work. 



SHUNT MOTOR ON ALTERNATING CURRENT 

59. A machine constructed in the same way as a direct- 
current shunt-wound motor, but provided with a laminated 
field, will not operate well on alternating current. This is 



Digitized by VjOOQ IC 



68 ALTERNATING-CURRENT APPARATUS § 19 

because the shunt field has a very high inductance, and the 
current in it lags nearly one-quarter of a cycle behind the 
E. M. F. It is thus very much out of phase with the arma- 
ture current, and but little power is delivered at the pulley. 
Such a motor would have a very low power factor, and 
while its performance might be improved by using a con- 
denser in connection with the field, it is a type that has 
never come into use in competition with the single-phase 
induction motor. 



REPrXSIOK MOTOR 

60. Fig. 38 illustrates the principle of a type of single- 
phase alternating-current motor known as a repulsion 

motor. The laminated 



field A A is excited by 
single-phase current. C is 
an armature provided with 
an ordinary direct-current 
winding connected to a com- 
mutator; d and e are thick 
brushes with their center line 
at an angle of about 45° with 
the center line of the poles. 
Suppose for the present that 
the brushes are not in con- 
tact with the commutator. 
When the field is excited, 
opposing E. M. F.'s will be 
induced in the two sides of 
the ring, like in a direct-cur- 
rent armature, and no cur- 
rent will be set up in the 
windings; no turning effort 
will be exerted, and the 
armature will not start. If 
a coil at g were short-cir- 
cuited by means of a brush, 




FlO. 38 



no current would be set up in it because the plane of 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 59 

this coil is in the same plane as that of the field; conse- 
quently, no torque would be exerted on the coil. If coil h 
were short-circuited, a heavy current would be set up in 
it because the alternating flux threads through it. How- 
ever, there is no field, from the pole pieces, at h to react on 
the induced current; hence, no torque will be produced. 
On the coils located between h and g a varying torque 
would be exerted, the maximum occurring in those coils 
situated about half way between the two extremes. If, 
therefore, two thick brushes rf, e are arranged so as to 
short-circuit a number of the coils lying in this region, a 
repulsive force will be exerted on the coils so short-circuited, 
and the armature will revolve. In Fig. 38 the short-cir- 
cuited coils are shown by the heavy lines. It will be noted 
that only those coils that are short-circuited by the brush 
are effective in producing rotation, so that comparatively 
few of the armature coils are utilized. More coils can be 
utilized and the repulsive effect made stronger by connect- 
ing the brushes together, as shown by the heavy dotted 
connection /. The repulsion motor is a special type of 
single-phase induction motor, because the currents in the 
armature are set up by the inductive action of the field and 
are not supplied from an outside source. 



WAGNBR STNGLK-PHASE INDUCTION MOTOR 

61, The Wagner single-phase induction motor has 
found extensive application because it gives a good starting 
effort with moderate current, and, while operating on 
single-phase circuits, compares favorably with polyphase 
induction motors. The general appearance of the motor is 
shown in Fig. 39. Fig. 40 is a sectional view showing the 
construction of the machine, particularly that part used 
when starting. The machine starts up as a repulsion 
motor, but after it has attained speed it operates as a 
regular single-phase induction motor with short-circuited 
armature. The field A^ Fig. 40, is provided with a simple 

44—37 



Digitized by VjOOQ IC 



60 ALTERNATINGM:URRENT apparatus § 19 

single-phase winding distributed in slots. The armature is 
provided with a regular direct-current winding placed in 
slots and connected to a commutator Fig. 41. This com- 
mutator is of the radial type and brushes are arranged to 
press against it, as shown. In Fig. 41 one complete ele- 
ment of the armature winding is shown, the field being 
wound for four poles. In this case a wave winding is 
represented so that two brushes only are required for a 




Pig. 30 



four-pole machine. The brushes are connected together 
as shown, so that when current is sent through the field, 
the motor can start up because of the repulsive action 
on the armature coils, as previously explained. When 
the motor has attained full speed, the commutator is 
short-circuited by means of a ring that connects all the 
segments together and at the same time the brushes are 
lifted from the commutator, thus preventing brush friction 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 61 

and wear except during the time that the commutator is 
actually in use. 



g 



The method of starting will be understood by referring 
to the sectional view, Fig. 40. The field laminations A are 



Digitized by VjOOQIC 



62 ALTERNATING-CURRENT APPARATUS § 19 

held in a cast-iron frame designed to protect the windings 
as much as possible. The armature windings are carried 
in slots in the core B, and these windings are attached to 
the bars c of the radial commutator. The brushes, repre- 
sented by the outlines d^ d press on these bars when the 
motor is at rest or running below full speed. The continu- 
ous ring that short-circuits the commutator is shown dX e^e\ 



PIG. 41 

this ring is mounted on the sliding sleeve /, the movements 
of which are controlled by the springy and the weights «/, w. 
When the motor attains full speed, the weights fly out on 
account of centrifugal force, and take up the position shown 
by w\ This movement forces sleeve /along the shaft and 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 



63 



brings e in contact with the commutator bars; at the 
same time the projection h pushes the brush-holder 
yoke to the left, and thus raises the brushes from the 
commutator. When the machine is shut down, the weights 
drop back and the spring g forces the collar to the right, 
thus removing the short-circuiting ring and bringing the 
brushes into contact with the commutator ready for the 
next start. 

62, Fig. 42 shows characteristic curves of a 15 H. P. 
motor. It will be noticed that at full load the power factor 



t.— 




















— 




























































































































Is 

























t_ 




«_ 


,^ 


















^ 










y^ 






^ 


^ 


















"^ 












^ 








/ 




J\ 


r 






































A 


/ 




A 


>^ 
































1 








1/ 




4 


^» 






































il 


1 


/ 


X" 




































^0 






/ 












































/ 


/ 














































1/ 


































^ 


X 












f 






























^ 


^ 










































^ 


^ 










«i 




























riT* 


'*^ 


C^ 
















:2 


























^ 






















^ 












































^ 














, 


^ 




"^ 






















^ 




































*i 


P, 




-^ 


























__ 


.I.. 




— 












J 


roji 


»w 


rot 


'KM 












f 










i 


^ 


i 


1 


i 








i 


• 


1 


9 


t 


i 


i 


4 


1 


• 


# 


• 







Pig. 48 



is 82 per cent., and the slip less than 2.6 per cent. This 
compares quite favorably with polyphase induction motors of 
corresponding output. 



Digitized by VjOOQIC 



«4 ALTERNATING^URRENT APPARATUS § 19 

ROTARY CONVERTERS 

63. It is often necessary to change direct carrent to alter- 
nating, and vice versa, and machines for accomph'shing this 
are known as rotary converters, or rotary transformers. 
These machines are also frequently referred to as synchro- 
nous converters, for reasons that will appear later. The 
transformation might be effected by having an alternating- 
current motor coupled to a direct-current generator, simply 
using the alternating current to drive the generator. An 
arrangement of two machines is, however, not usually neces- 
sary, although such motor-generator sets are used to some 
extent. Rotary converters are largely used for changing alter- 
nating current to direct for the operation of street railways, 
electrolytic plants, etc. 

SINGLE-PHASE CONVERTERS 

64. Suppose an ordinary Gramme ring armature to be 




Fig. 48 



revolved in a two-pole field, as shown in Fig. 43; a direct 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 65 

current will be obtained by attaching a circuit to the 
brushes a, a\ If, instead of the commutator, two collector 
rings were attached to opposite points of the winding, an 
alternating current would be obtained in a circuit connected 
to ^, b\ If the machine be equipped with both commutator 
and collector rings, the armature may be revolved by means 
of direct current led in at the brushes ^, a\ thus running it 
as a motor instead of driving it by a belt. The conductors 
on the revolving armature will be cutting lines of force 
.just as much as they were when the machine was driven by 
a belt; therefore, an alternating current will be obtained 
from the rings ^, b\ In other words, the machine acts as a 
converter, changing the direct current into a single-phase 
alternating current. If the operation be reversed and the 
machine be run as a synchronous alternating-current motor, 
the alternating current will be transformed to direct. 

65. In the above single-phase rotary converter, it is evi- 
dent that the maximum value of the alternating E. M. F. 
occurs when the points i, 1\ to which the rings are con- 
nected, are directly under the brushes a^ a! \ that is, the 
maximum value of the alternating E. M. F. is equal to the 
continuous E. M. F. For example, if the continuous 
E. M. F. were 100 volts, the effective volts on the alterna- 

tinc:-current side would be — = = 70.7 volts. Therefore, 

if E is the alternating voltage and V the direct, we may 
write for a single-phase rotary converter, 

E = .707 V , (7) 



TWO-PHASE CONVERTEBS 



66, By connecting four equidistant jpoints of the wind- 
ing ^, ^, d^ and e^ Fig. 44, to four collector rings, we would 
have a two-pole two-phase, or quarter-phase, converter. In 



Digitized by VjOOQIC 



66 ALTERNATING-CURRENT APPARATUS § 19 
this case we would have two pairs of lines leading from the 




PIO. 44 



brushes i, 1\ 2, 2'. The E. M. F. between 1 and 1' or 
between 2 and 2' would be given by formula 7. 



THREE-PHASE CONVERTERS 

67, By connecting three equidistant points as shown at 
^, c, and d, Fig. 45, a three-phase converter is obtained. 
Since all direct-current armatures have closed-circuit wind- 
ings, it follows that the connections on the alternating- 
current side of a three-phase rotary converter are always A, 
the Y connection not being possible. If E be the effective 
voltage between the lines on the alternating side of a three- 
phase rotary converter and V the voltage of the direct- 
current side, 

£=.612 F (8) 

If such a converter were supplied with direct current at 
100 volts pressure, alternating current at 61.2 volts would be 



Digitized by VjOOQIC 



§ 19 ALTERNATING-CURRENT APPARATUS 67 
obtained ; and if it were desired to obtain 100 volts direct 




PIO. 46 

current from alternating-, the alternating side would have to 

be supplied at a pressure of 61.2 volts. 

The proof of the above relation is as follows : 

Suppose the closed-circuit winding of a two-pole rotary 

transformer is represented in Fig. 46. 




Pig. 40 



The E. M. F. V obtained across the diameter of the wind- 
ing would be the continuous E. M. F. of the machine. It 



Digitized by VjOOQ IC 




Digitized by VjOOQIC 



g 19 ALTERNATING-CURRENT APPARATUS 69 

would also be the maximum alternating E. M. F. for a 
single-phase rotary transformer. For a three-phase rotary, 
the winding would be tapped at three equidistant points, 
a^ b, and r, and the maximum values of the alternating 
E. M. F. between the three collector rings would be repre- 
sented by the three lines a b^ b c^ and c a. To obtain the 
value of this E. M. F. in terms of the continuous E. M. F., 
from the center d draw the line d e perpendicular to a b. 
Then the angle a de = 60*". 

ae = ^ Vsm 60'' 
^ * = 2 rt ^ = F sin 60° 

a b ^ maximum E. M. F. of alternating-current side of 
machine. Then, the 

effective- E. M. F. i? = .707 Fsin 60° 

= .707 X .866 X V 
= .612 V 



MUIiTIPOT^AR ROTARY CONVERTERS 

68. The windings shown in Figs. 43, 44, and 45 show the 
connections for two-pole machines; but rotary transformers 
are nearly always made multipolar in order to reduce the 
speed of rotation. In the single-phase machine shown in 
Fig. 43, it was necessary to have only one connection to 
each ring; in a multipolar machine it is necessary to have as 
many connections to each ring as there are pairs of poles on 
the machine. Fig. 47 shows the connections for a six-pole 
single-phase converter. Here the ring / is connected to the 
points gy //, and/, while ring 2 is connected at r, d^ and ^, 
these points being the equivalent of 180° apart. If only two 
connections were made, as in Fig. 43, the whole winding 
would not be utilized. Fig. 48 represents the same armature 
connected up as a three-phase converter. Here each of the 
three rings has three connections, as before, and these con- 
nections are the equivalent of 120° apart. For example, the 
angular distance from ^ to ^ is one-third the distance from 



Digitized by VjOOQ IC 



70 ALTERNATING-CURRENT APPARATUS § 19 

north pole 'to north pole, which represents 360**. Such a 
winding would, therefore, have the three connections r, d^ e 
for ring /; fyg^h for ring 2\ and k^ /, m for ring 3, there 
being as many connections for each ring as there are pairs of 
poles. 

69. Fig. 49 shows the general appearance of a three- 
phase rotary converter. The three collector rings are 




PIO. 49 



seen at the left-hand end of the machine, and the commu- 
tator is shown at the right. Like alternators, rotary con- 
verters are built for a large range of output and frequency. 
In the foregoing explanations regarding the rotary con- 
verter, simple ring windings have been used, because of the 
ease in following the connections. In converters as actually 
built, the windings are of the toothed-drum type and in 
fact are the same as used on direct-current multipolar 



Digitized by VjOOQIC 



§ 19 ALTERNATING-CURRENT APPARATUS 71 

dynamos. Either series- or parallel-wound drums may be 
used. With a series-wound drum, only one connection to 
each collector ring is necessary, no matter how many poles 
the machine may have. With a parallel-wound drum, there 
would be a connection to each ring for each pair of poles, as 
shown for the ring windings given in the preceding article. 



OPERATION OF ROTARY CONVERTERS 

70, Rotary converters are used to change alternating 
current to direct current in the great majority of cases. 
They are used comparatively seldom to change direct- 
current to alternating, though special cases sometimes arise 
where it is advantageous to use them in this way. When 
used to change direct current to alternating, they are fre- 
quently called inverted rotaries, though there seems to be 
little need for this special name, because the machine itself is 
in nowise different; the only difference is in the manner in 
which it is used. 

A r6tary converter, run in the ordinary way from alter- 
nating-current mains, operates as a synchronous motor, 
and hence runs at a constant speed. Its speed will not vary, 
no matter what load may be taken from the direct-current 
side, provided the speed of the alternator that supplies the 
current is maintained at a constant value. 

71, Heatingr of Rotary Converters. — While the rotary 
converter behaves on the one hand like a synchronous motor 
and on the other like a direct-current dynamo, it has a 
number of peculiarities due to the combination of these two 
in a single piece of apparatus. One of the chief differences 
is in the heating effect. If a rotary converter is run as a 
direct-current dynamo, the output that can be obtained 
from it is considerably less than if it is run as a converter; 
that is, with the same heating effect in the armature, a given 
machine is capable of a greater output when run as a con- 
verter than when operated as a direct-current dynamo. 



Digitized by VjOOQ IC 



72 ALTERNATING-CURRENT APPARATUS § 19 

Thi^ is not true of the single-phase converter, but these 
are seldom if ever used in practice. The reason that a 
multiphase converter gives a greater output than a direct- 
current dynamo of the same size may be explained in a 
general way as follows. 

72, The currents that actually flow in the armature 
conductors are of peculiar character. They are the differ- 
ence between the alternating current supplied and the 
direct current given out. For example, take the simple 




Pig. 60 

case of the single-phase converter shown in Fig. 50, where 
the various coils are shown connected to the commutator; 
c and d are the brushes from which direct current is 
delivered, and g, h the brushes through which alternating 
current is supplied. At the instant a coil passes under c 
or </ the direct current in it is reversed, but while it is turn- 
ing through the half revolution between c and d the direct 
current remains at a constant value. Take the coil ^, for 
example, this being the coil situated midway between the 
taps ^, b and the brushes c^ d. The direct current in this 
coil may be represented by the flat-topped wave a-b--c-d~e^ 
etc., Fig. 51 {a). The alternating current in the half of the 
armature from a around to d, including coil e, will be a 



Digitized by VjOOQIC 



§ 19 ALTERNATING-CURRENT APPARATUS 73 

maximum when the armature is in the position shown, and 
will have become zero when coil e comes under brush c. In 
other words, the alternating current in coil e will reverse at 
the same instant as the direct current, and hence may be 
represented by the wave m-n-o-p^ etc., which passes through 
zero at the same instant as the flat-topped wave. The 



*\ 




« » /^"^ 


* /- 




\ 




{ N 


•• 


/ 


t 


V / 


„ _ ^ . 


r y 




• 


1 


' V ^ / • 


w 


1 ■ " 



w 



X 


k /] 


^,-v,^ 


K / 


Z' 


N 




V N 




V 



PIO. SI 

alternating current and direct current flow in opposite 
directions, because in one case the machine acts as a 
dynamo and in the other as a motor. 

The current that actually flows in the conductors is found 
by combining the two curves shown in (a)^ thus giving the 
resultant curve shown in (b). This peculiar wave repre- 
sents the current in coil e\ it is easily seen that the average 
value of the current is much smaller than that represented 
by the flat-topped wave in (^), which represents the current 
that would flow in the coil if the machine were used as a 
direct-current dynamo; hence, the heating will be less 
because the heating is proportional to the square of the 
effective value of the current. 

73. It should be noted that the curves given in Fig. 51 
represent the component and resultant currents in coil e 



Digitized by VjOOQIC 



74 ALTERNATING-CURRENT APPARATUS § 1^ 

only. In the case of a coil situated on either side of e^ the 
direct-current or flat-topped wave would not pass through 
zero at the same time as the alternating current, and the 
nearer the coil is to the tapping-in points, the greater differ- 
ence in phase there would be between the two curves. The 
resultant current in each of the coils is therefore different, 
and a different heating effect is produced, the coils near the 
tapping points heating considerably more than the coils 
situated near e. As the number of tapping points and the 
number of phases is increased, the more uniform does the 
heating effect become, and the less also it becomes as com- 
pared with that produced when the machine is run as a 
direct-current generator. In a single-phase rotary the sum 
of the heating effects in all the coils is greater than if the 
machine were run as a direct-current dynamo, because the 
maximum value that the resultant current reaches in the 
coils near the tapping-in points is very much higher than 
the value of the direct current. The heating in these coils 
is therefore excessive, and brings up the total heating effect 
to an amount greater than that of the corresponding direct- 
current machine. As the number of phases is increased, 
and the angular distance between the tapping points 
decreased, the total heating effect is decreased, and for 
multiphase converters is considerably less than that of the 
corresponding direct-current machine. 

74, This difference in the heating effect of a rotary 
converter as compared with a direct-current generator is 
important because it means that such a machine can be 
made smaller and lighter for the same output than a direct- 
current dynamo. A machine connected up for three phases 
will give a larger output than the same machine connected 
for single-phase operation, and a six-phase machine will 
give a still larger output with the same amount of heating. 
A six-phase machine is one in which the windings are 
tapped at six equidistant points and the terminals led to six 
collector rings. Such a machine would, be supplied by six 
currents differing in phase by 60° ; these currents are easily 



Digitized by VjOOQ IC 



c 

^ 



§ 19 ALTERNATING-CURRENT APPARATUS 75 

obtained from a three-phase system by a special arrange- 
ment of transformers. 

Six-phase converters are used to a considerable extent in 
practice because of the increased capacity obtained, but a 
larger number of phases has not been employed to any 
extent, because the gain in output is more than offset by 
the complexity caused by the increased number of collector 
rings, brushes, transformers, and transformer connections. 
If the output when used as a direct-current machine is taken 
as 1, then the output when run as a three-phase converter 
will be 1.3-1:; as a quarter-phase, or four-phase, converter, 
1.G4; and as a six-phase converter, 1.96. In other words, 
the output as a three-phase converter is about one-third 
greater than as a direct-current machine, and the output as 
a six-phase converter is almost twice as great. The above 
figures assume that the armature current and E. M. F. are 
in phase. If there is a lag between the current and 
E. M. F., the output when operated as a converter is less 
than that represented by the above figures. Six-phase con- 
verters and the transformer connections used with them 
will be taken up later. 

75. Voltage Regrulatlon of Rotary Converters. — It 

has already been noted that the rotary converter, as 
ordinarily used, operates as a synchronous motor, and that 
the ratio of the alternating voltage to that on the direct- 
current side is a fixed quantity. For example, suppose it 
were desired to transform alternating current at 2,000 volts 
to direct current at 500 volts, suitable for operating a street 
railway. We will suppose that a three-phase rotary trans- 
former is employed. Then it follows that the alternating 
current must be supplied to the machine at a pressure 
oi E=i .612 V = .612 X 500 = 306 volts. The alternating 
current would, therefore, be first sent through static trans- 
formers wound so as to reduce the pressure from 2,000 to 
306 volts, and the secondary coils of these transformers 
would be connected to the alternating-current side of the 
converter. 



Digitized by VjOOQIC 



76 ALTERNATING-CURRENT APPARATUS § 19 

If it is necessary to vary the direct E. M. F. through 
any considerable range, the alternating E. M. F. must be 
varied also. This is frequently accomplished by using trans- 
formers provided with a number of taps from the secondary 
windings, so that the turns can be cut in or out, thus 
varying the alternating E. M. F. applied to the rotary. 
Another plan is to insert potential regulators between the 
secondary of the transformers and the collector rings of the 
rotary. Both these methods of regulation will be described 
more fully in connection with the uses of rotary converters 
in power-transmission systems. 

76. If the field excitation of a rotary converter is 
changed, the current taken by the machine can be made 
lead the E. M. F. or lag behind it, according as the field is 
strengthened or weakened from the amount corresponding 
to a power factor of unity. Converters are nearly always 
operated with a field excitation that will give a power 
factor as near unity as possible, because then, for a given 
load on the direct-current side, they take the minimum 
current from the line. Changes in the field magnetization 
will cause changes in the direct-current voltage through a 
limited range. Increasing the excitation makes the machine 
have the same effect as a condenser, and causes the voltage 
at the terminals to rise, thus increasing the direct E. M. F. 
Lowering the E. M. F. makes the current lag and lowers 
the applied E. M. F. The range of regulation that can be 
obtained by this method is, however, comparatively small, 
and besides, it has the disadvantage of decreasing the 
power factor, so that wherever any considerable range is 
desired the applied E. M. F. is varied, as described above. 

77. C>i>eration From Dlrect-Curreut Side. — When a 
rotary is supplied with direct current, it runs as a direct- 
current motor and its behavior under certain circumstances 
is quite different from that when it is operated from the 
alternating-current side. Weakening the field will cause 
the armature to speed up, and strengthening the field will 



Digitized by VjOOQ IC 



§ 19 ALTERNATING-CURRENT APPARATUS 77 

slow it down as with any direct-current motor. Altering 
the field strength will not change the alternating E. M. F., 
because every change in field strength is accompanied by a 
change in speed, and the alternating E. M. F. that is gen- 
erated in the conductors remains the same. In order to 
vary the alternating E. M. F. it is necessary to vary the 
direct E. M. F. applied to the armature, or else use potential 
regulators in the lines leading from the alternating side of 
the rotary; the latter would in most cases be the more 
practicable method. 

There is a peculiar effect sometimes met with when rota- 
ries are operated with direct current. If the alternating- 
current side is accidentally short-circuited, the machine is 
apt to race, because the heavy lagging current set up in the 
armature exerts a powerful demagnetizing action on the 
field and weakens it to such an extent that the machine 
attains a high speed. 



DOUBLlE-CUBRBNT GENBRATOBS ' 

78. If a machine constructed in the same way as a rotary 
converter be driven from some outside source of power, it 
will deliver direct current from the commutator end and 
alternating current from the collector rings. A machine so 
operated is called a double -current generator. The full out- 
put of the machine may be delivered as direct current, as 
alternating current, or partly as one and partly as the other, 
provided, of course, that the combined output on the two 
sides does not exceed the capacity of the machine. 

Double-current generators are useful where it is desired 
to have part of the output of a plant as direct current for 
utilization near at hand, and part as alternating current for 
transmission to distant points. 

While the construction of a double-current generator is 
similar to that of a rotary converter, the heating action in 
the armature is quite different. In the rotary converter, we 
have just seen that the current in the armature conductors 
is the difference between the current taken as a motor and 



Digitized by VjOOQ IC 



78 ALTERNATING-CURRENT APPARATUS § 19 

the current delivered as a dynamo. In the double-current 
generator, the action is that of a generator on both sides, 
and the currents in the armature conductors are the sum of 
the currents delivered to the two sides. The heating is, 
therefore, considerably greater than if the same machine 
were used as a liDtary converter delivering an equal output. 
Another point of difference in the action of the two machines 
lies in the effect of armature reaction. In a polyphase rotary 
converter, the armature reaction due to the direct current is 
offset by that due to the alternating current flowing in the 
opposite direction. In the double-current generator both 
currents react on the field and produce the same effects as 
the armature reaction in a direct-current generator. In 
case a heavy inductive load be thrown on the alternating- 
current side, the lagging current produces a powerful 
demagnetizing action on the field. This would greatly 
affect the voltage if the machine were self-exciting, as with 
ordinary direct-current machines, and in order to secure a 
stable field, double-current machines are generally separately 
excited. In most cases they are also provided with a series- 
field winding, through which the direct current passes 
and strengthens the field as the load increases. 



Digitized by VjOOQIC 



INDEX 



NoTB.— All items in this index refer first to the section and then to the pasre of the 
section. Thns "Alternation 16 4" means 
section 16. 



Sec. Page 



Accnmnlatively wound motors . . 15 

Action of motor 15 

" " shunt motor ...... 15 

** ** the armature 12 

Addition of sine curves 10 

Airsrap 14 

" •• Density in the 14 

Altematinff-current apparatus . . 19 
" " circuits. Power 



expended in 

generator . . 

lines. Drop in 

measurintr in- 
struments . 

motors .... 

Series motor 
on 

Shunt motor 
on 

voltmeter, 
Wasmer . . 

voltmeter, 
Weston . . . 



17 

17 

•• currents 16 

17 

Alternation 16 

Alternator. Monocyclic 18 

** Westlnifhouse two- 
phase, compound- 
wound 

•• Westinsrhouse three 
phase, compound 
wound 



Alternators 



Calculation of B. M. 
P. firenerated by . . 
Construction of . . . 
Field excitation of . . 
Polyphase 18 



31 
2 

20 
3 
9 

25 
6 
1 

la 

5 
32 

34 
23 



19 57 



19 57 



46 

1 

1 

4 

51 



18 53 



18 


55 


18 


X 


18 


12 


18 


14 


18 


5 


18 


20 


18 


32 



that alternation will be found on page 4 of 

Sec. Page 
Alternators, Revolvinir-field and 

inductor 18 25 

** Slnffle- phase 18 1 

•* 18 32 

Three-phase 18 39 

Two-phase 18 32 

** with closed-circuit 

armature windlnirs 18 52 
Ammeter, Series-transformer for 19 20 
■ Ammeters and voltmeters. Hot- 
wire 17 85 

Ampere conductors per Inch of 

circumference 14 6 

** turns, Calculation of 

cross 14 33 

" ** Compensation for 

cross .... 
•• " Cross and back 
•* " Effect of cross 
•• " on field. Determi- 
nation of .... 13 83 
•• "to offset armature 

reaction .... 14 36 

AnsrleoflafiT 16 42 

Apparatus. Altematinsr-current . . 19 1 

Auxiliary 15 35 

Area. Brush-contact 14 22 

Armature 12 2 

15 80 

Action of the 12 3 

" and commutator. De- 

sifimof 14 43 

Barrel-wound 12 24 

coil, short-circuited or 

reversed 15 91 

coils, Locatinsr short- 
circuited 15 99 

conductors 14 12 

14 9 

Insulation of 14 12 



13 


46 


14 


34 



Digitized by VjOOQ IC 



INDEX 







Sec. 


Paze 


Armature. Conttrtiction of the . . 


13 


50 


•♦ 


core 


12 
18 


8 


" 




50 


" 


•• 


18 


79 


** 


" and windinsr, I>e- 








siffn of 


14 


1 


•• 


" Lensrth of the . . . 


14 


4 


** 


*' losses 


12 


18 


•• 


cores slotted 


12 


20 


'* 


Diameter of the .... 


14 


8 


•• 


Dimensions of 


14 


9 


•* 


flux 


13 


82 


•* 


•• 


14 
13 


81 


•• 


HeatiuflTof 


85 


•• 


/» R loss 


14 


68 


•• 


losses and heatins: . . . 


18 


70 


•« 


Open-circuited 


15 


91 


(« 


or field, Wronsr connec- 








tion of 


15 


85 


•• 


reaction 


IS 


42 


•• 
•• 


" Ampere* turns 


15 


17 




to offset . . . 


14 


86 


•• 


•* Effects of . . . 


18 


49 


•» 


•* " ** . , . 


14 


88 


•• 


resistance. Estimation 








of 


13 


70 


•* 


shaft. Sprunsr 


15 


90 


*• 


slots 


18 


58 


•• 


spiders 


18 


52 


" 


Squirrel-casre 


19 


35 


•* 


Startinar with resistance 








in 


19 


46 


•• 


surface covered by poles 


14 


8 


*• 


teeth 


13 


80 


*• 


Tests for defects in . . 


15 


97 


•' 


w i n d i n if s. Alternators 








with closed-circuit . . 


18 


52 


•* 


windinsrs. Closed-coil . 


12 


20 


*• 


Open-coil . . 


18 


15 


Armatures. Cross-connection of . 


13 


18 


" 


Cylindrical-wound . . 


12 


24 


•• 


Drum and rinir .... 


12 


10 


•• 


for induction motors . 


19 


34 


«« 


Heatinsr of 


15 


80 


*• 


Overloaded 


15 


81 


*• 


Perforated 


12 


20 


** 


Smooth-core 


12 


20 


*• 


Smooth-core and tooth 


13 


57 


" 


Toothed ....... 


12 


18 


•* 


•• 


12 


20 


Arrantrement and size of conductor 


14 


20 


Automatic release. Rheostat with 


15 


88 


*• 


" startintr rheo- 








stats. Types 








of 


15 


44 



Sec. Page 

Automatic starting rheostats ... 15 56 

Autotransformer 19 19 

** or compensator. 

StartiniT .... 19 41 

Auxiliary apparatus 15 85 

B 

Back ampere-turns and cross am- 
pere-turns 13 46 

Bar. Hiffh or low 15 90 

'• to-bartest 15 97 

Bars. Higrh 15 79 

'* Low 15 79 

Barrel-wound armature 12 24 

Bearintr friction and windage ... 14 68 

Bearinsrs 13 64 

" and bedplate 14 62 

Low 15 90 

Bedplate and bearinsrs 14 62 

Belt-driven and direct-driven 

dynamos 12 15 

•* slippinsr 15 92 

Bipolar dynamos 12 15 

Booster-teaser system of control . 15 65 
Brown St Sharpe srausre. Maffnet 

wire 13 72 

Brush constant-current dynamo . 13 18 

•* contact area 14 22 

•* friction. Commutator ... 14 68 

Hiifh-resistance 15 91 

" holders 12 18 

and rocker 14 58 

Brushes 12 5 

15 76 

" Improperly set 15 89 

•' making: poor contact . . 15 90 

Position of 12 27 

Radial 15 76 

Tangential 15 76 

Thickness of 14 24 

*' Use of hiffh-resistance . . 13 39 

Buckins: 18 12 

C 

Calculation heatins: 14 17 

" of cross ampere^tums 14 S3 

*• " E. M. F. and power 18 27 

•• E. M. F. fi:enerated 

by alternators . . 18 14 
*' •' field winding for 115- 

125 volts 14 87 

"flux 14 14 

" •* power 13 28 

** " power expended in 

alternatinsr-cur- 

rent circuits ... 17 18 



Digitized by VjOOQ IC 



INDEX 



XI 



Sec. 
Capacity and resistance. Circuits 

containincr .... 17 
'* " resistance in par- 
allel 17 

" " self-induction . . . 1« 
" ** self-induction. Cir- 
cuits containing . 17 
'* " ^elf-induction in par- 
allel 17 

'* Circuits containinff ... 16 

Line 17 

" reactance 16 

•• resistance, and self- 
induction 17 

** resistance and self-induc- 
tion. Circuits contain- 

inir 17 

** resistance and self-induc- 
tion in parallel .... 17 
Care of dynamos and motors ... 15 

** machines 16 

Characteristics of compound- 

wound machines 12 

" of series-machine 12 

*' shunt machine 12 

** " the dynamo . . 12 

Char^res, Condenser . r 16 

Circuit. Form of masrnetic .... 12 

Masmetic 12 

14 

14 

•* Power fagtor of a 17 

** Series motor on constant- 
current 16 

** Series motor on constant- 
potential 16 

Circuits. Calculation of power in 

alternatinsr-current . . 17 
** containing capacity ... 16 
** " resistance . . 16 
•* ** resistance 
and capac- 
ity 17 

•• *• resistance 
and self-in- 
duction . . 16 
** containing: resistance and 
self-induction and 

capacity 17 

* containins: self-induction 16 
** containins: self-induction 

and capacity 17 

in field. Short 16 

Short . . . . : 16 

Classes of instruments ...... 17 

" ** motors 16 



Pare 


Closed-circuit armature windings. 


Sec. 


PO£i 


1 


Alternators with 


18 
12 


62 




Closed-coil and open-coil windings 


11 


6 


** armature windings . . 


12 


20 


28 


" windincrs 


13 


1 




Coefficient of magnetic leakagre . . 


13 


82 


6 


Coils and slots. Numt>er of ... . 


14 


11 




" Construction of 


13 


60 


7 


" Loss in shunt 


14 


69 


43 


" Manner of windinfir the . . . 


12 


21 


82 


** Pitch or spread of 


12 


21 


48 


** to commutator. Connecting 


12 


26 




Collector rin^s 


12 


5 


1 


Commercial efficiency 


13 


31 




•* ** 


16 


8 




of motors . 


16 


10 


8 


Commntatinff frinsre 


13 


89 




Commutation 


12 


13 


16 


and sparkinff .... 


13 


36 


76 


*' Requirements for 






76 


sparkless 


13 


41 




Commutator 


12 


9 


67 


•• 


16 


78 


60 


and armature. De- 






63 


sifim of . . - 


14 


48 


66 


** brush friction .... 


14 


68 


44 


Connecting coils to 






46 


the 


12 


26 


42 


*' Design of 


14 


21 


23 


Dirty 


16 


90 


28 


•• Rouffh or eccentric . 


15 


89 


26 


" sesrments. Number of 


14 


10 




Worn 


16 


90 


29 


Commutators 


13 


64 




Compensated wattmeter . ... 


17 


60 


24 


Compensator or autotransformer. 








Startintr 


19 


41 


18 


" Stanley startintr . . 


19 


41 


43 


Components of applied E. M. P. . 


16 


88 


80 


Wattless and power 
Composition and resolution of cur- 


17 


26 




rents and B. M. P.'s 


16 


20 


1 


Compound dynamo converted to 








• motor 


16 


35 




•* windinar 


12 


66 


88 


" wound alternator. 
We^tinsrhouse three- 








phase 


18 


56 


8 


wound alternator. 




• 


31 


Westinfifhouse two- 








phase 


18 


58 


6 


'* wound machines. 






83 


Characteristics of . 


12 


67 


86 


wound motors . . . . 


16 


30 


36 


Computation of field windings . . 


14 


29 


19 


Condenser charges 


16 


44 



Digitized by VjOOQ IC 



Xll 



INDEX 





Sec. 


Pazt 


Condenser, E. M. P 


16 


45 


Conductor. Armature 


14 


12 


Size and arransrement 






of 


14 


20 


Conductors, ampere, per inch of 






circumference . . 


14 


6 


" Armature 


14 


9 


" Inductors or face . . 


12 


10 


Insulation of arma 






ture 


14 


12 


Connecting coils to the commu- 






tator 


12 


26 


shunt coils, Methods 






of 


16 


40 


Connection of field or armature. 




Wrontr 


15 


85 


" shunt field, Wronff 


15 


43 


Connections 


14 


64 


Field 


19 


49 


of shunt field. Wrongr 


15 


88 


Series motor 


15 


60 


" Shunt motor 


15 


37 


Star and delta .... 


18 


42 


Consequent poles 


12 


46 


Constant-current circuit. Series 






motor on . . 


15 


29 


dynamo, Brush 


13 


18 


** " dynamo, Thom- 






son-Houston . 


13 


19 


** potential circuit. Series 






motor on . . 


15 


24 


** " dynamos . . . 


13 


34 


•* ** dynamos. Out- 






put of ... . 


18 


34 


mains. Series 






motor across 


15 


27 


Construction of alternators .... 


18 


5 


•• coils 


13 


60 


'• ** core and spider . 


13 


60 


" •• dynamo 


12 


1 


** ** field frame and 






field coils . . . 


14 


51 


" ** the armature . . 


13 


50 


** ** transformers . . 


19 


14 


Control. Booster-teaser system of 


15 


65 


by variation of field reluc- 






tance 


15 


67 


Field 


IS 


23 


'* Multivoltaire speed .... 


15 


59 


Rheostatic 


15 


22 


Teaser system of .... 


15 


65 


Conversion. KflRcicncy of 


15 


8 


of mechanical into 






electrical cnersry . . 


13 


32 


Converters, Heatintr of rotary . . 


19 


71 


*' Multipolar rotary . . 


19 


69 



See. 
Converters, Operation of rotary . 19 

Rotary 19 

Sinffle-phase 19 

** Synchronous 19 

" Three-phase 19 

'• Two-phase 19 

** Voltaire refTulation of 

rotary 19 

Copper wire. Reactance and im- 
pedance 17 

Core and spider. Construction of . 13 
** " windin^r. Desism of arma- 
ture 14 

" Armature 12 

13 

13 

** Diameter of 14 

** losses, BflFect of 19 

" transformers 19 

Cores, Masmet 13 

14 

Counter E. M. F. of motor .... 15 
Cross ampere- turns and back 

ampere-turns 13 

** ampere -turns. Calculation 

of 14 

" ampere -turns, Compensa- 
tion for 14 

ampere-turns. Effect of . . 14 

*• connecting: rings 14 

** connection of armatures . . 13 

" Flyinsr 15 

Current and torque. Relation 

between 16 

B. M. P.. and output. Re- 
lation between IS 

** MasrnetiziniT 19 

" Series motor on altcma- 

tinir 19 

** Shunt motor on altema- 

tinsr 19 

Currents, Altematinir 16 

17 

'* andE.M.P.'s. Composi- 
tion and resolution of 16 
Curves of induction motor. Char- 
acteristic 19 

Cycle 16 

Cylindrical- wound armatures ... 12 

D 

Defects, Field-coil 15 

" in armature. Test for . . 15 

Delta and star connections .... 18 

Densities. Masmetic 13 

Density in teeth 14 



71 
64 
64 
64 
66 
65 

75 

SI 

dC 

1 
3 
50 
7S 
16 
12 
14 
87 
23 
3 



34 

34 

20 
13 
81 



16 

46 
6 

S7 

67 

1 
1 

20 

61 

4 
24 



82 
97 
42 
79 
14 



Digitized by VjOOQ IC 



r 



INDEX 



Xlll 



Sec. Page 

Density in the air trap 14 6 

'* of lines of force 12 44 

Dcsism, Mechanical 14 42 

15 73 

" of a lO-horsepower series 

motor 16 72 

" " ** 10-horsepower shunt 

motor 15 71 

lOO-kilowatt dynamo 14 1 

** " armature and commu- 
tator 14 48 

" " armature core and 

windintr 14 1 

•* ** commutator 14 21 

" ** direct-current motors . 15 68 

** the field masmet .... 13 77 

Determination of output 15 69 

Diameter and peripheral speed . . 14 21 

" of core 14 16 

" the armature .... 14 8 

DifTerentially wound motors ... 15 30 

Dimensions of armature 14 9 

'• slot 14 13 

Direct-current motors 16 1 

Desifirn of . 16 68 
" " " Operation 

of .... 15 1 
" ** side. Operation 

from 19 76 

*• driven and belt-driven 

dynamos 12 15 

Direction and speed of rotation . . 19 28 

Double-current generators .... 19 77 

parallel windinc: 13 11 

** ** " Sinely re- 
entrant 12 82 

** series windings 12 39 

18 11 

windinc: 12 80 

Sinsrly reentrant . 12 31 

Drop in altematingr-current lines . 17 82 

Drum and rinsr armatures .... 12 10 

windingfs 12 10 

13 4 

Ducts. Ventilating 13 57 

Dynamo 15 1 

and motor rotation ... 15 32 

'* Brush constant-current . 13 18 

Characteristics of the . . 12 56 

** Construction of 12 1 

•' design 12 1 

•• of a 100-kiIowatt . . 14 1 

" electric machine 12 1 

** Essential parts of a ... 12 2 

•• Failure of. to generate . . 15 84 

Self-excited 12 3 



Sec. Page 



Dynamo Separately excited ... 12 

shafts 18 

Theory of the 12 

" Thomson-Houston con- 
stant-current 13 

windings 12 

Dynamos and dynamo design . . 12 

. . 13 

. 14 



" motors. Care and 

operation of . . . 

** motors compared . 

Bipolar 

Classes of 

Constant-potential . . 
Direct-driven and belt- 
driven 

Electrical efficiencies of 
Output of constant- 
potential . 

Unipolar 



15 
15 
12 
12 
13 84 



£ 



Eddy-current loss 



lost 

Effect of core losses 

" ** magnetic leakage .... 

** resistance of primary 

and secondary coils . . 

Effects of self-induction 

Efficiencies of dynamos. Electrical 

Efficiency 

** Commercial 



Electrical 



" Motor 

" of conversion 

" " motors. Commercial 

Electric generator 

Electrical efficiencies of dynamos 

** efficiency 



" energy. Conversion of 
mechanical into . . . 

" losses 

•• resonance 

Electrodynamometers 

Electrostatic voltmeters 

" Stanley . 

E. M. F. and power. Calculation of 
" " " Components of applied . 

Condenser 

** ** " generated by alternators. 
Calculation of 



12 
14 
18 
19 
19 

19 
16 
14 
14 
13 
15 
18 
15 
15 
15 
15 
12 
14 
13 
16 

13 
13 
17 
17 
17 
17 
13 
16 
16 



3 

62 

1 

19 

41 

1 

1 

1 

75 
1 

15 
2 



15 
2 



34 
24 



18 
18 
74 
12 
11 

11 

29 
2 

67 

31 
8 

30 
8 
8 
8 

10 
1 
2 

30 
8 

82 
30 
13 
44 

54 
55 
37 
38 
45 



18 14 



Digitized by VjOOQ IC 



XIV 



INDEX 



Sit, Pagt 
B. M . P. of motor. Counter .... 15 8 
/ output and current. Re- 
lation between 18 46 

•• " •• Value of induced 16 88 

wave forms 16 1 

B. M. P.'ft and currents, Composi- 
tion and resolution of 16 20 

Bnertry. Conversion of mechanical 

into electrical 13 82 

Estimation of output 14 1 

Excitation of alternators, Field . . 18 20 

BxcitinfiT the field. Methods of ... 12 54 

F 

Face conductors or inductors ... 12 10 

Factor. Form 16 25 

Failure of dynamo to srenerate . . 15 84 

•* *' motor to start 16 88 

Faults. Testinsr for 15 92 

Field, Buildins: of the 12 69 

•* coil defects 15 82 

•• Short-circuited 15 94 

•' coils .- 12 18 

** and field frame .... 14 51 

*• Moisture in 15 88 

" ** opposed 15 87 

•• •* Testinc: for open -cir- 
cuited 15 93 

** connections 19 49 

*• control 15 28 

•* Determination of ampere- 
turns on 13 88 

** excitation of alternators . . 18 20 

'• flux 18 82 

** frame and field coils .... 14 51 

" masrnet 12 2 

Desigmofthe ... 13 77 
** masrnets. Multipolar .... 12 15 
" Method of exciting: the ... 12 54 
•• or armature. Wrong: connec- 
tion of .... 15 85 

•• reluctance. Control by vari- 
ation of 15 67 

•* Revolving 19 30 

*• Short circuits in 15 83 

" Weak 16 92 

*• winding: for 115-126 volts . . 14 37 

•• winding:s 18 84 

Computation of . 14 29 

** Wrong: connections of shunt 16 88 

Flux. Armature 13 82 

14 31 

•' Calculation of 14 14 

" Field 13 82 

" in pole pieces 14 23 

Plying: cross 15 81 



iSr. Pa£t 

Force. Density of lines of 12 44 

Form factor 16 25 

Formula 12. Values of K ..... 18 63 

Frames, Mas:net 13 81 

Frequency 16 4 

Friction and windag:e. Bearing: . . 14 68 

*• Commutator brush . . . J4 68 

Pring:*. CommutatiAg: 13 89 

G 

Generator, Altemating:-current . . 12 5 

Electric 12 1 

Induction 19 88 

125-volt 14 67 

Generators, Double-current .... 19 77 

250-volt and 600-volt . 14 70 

Grounds 16 92 

** between winding: and 

frame 15 05 

H 

Heating: and loss. Armature ... 18 70 

'• calculations 14 17 

** of armature . 13 85 

15 80 

" ** rotary converters ... 19 71 

Higrhbars 15 79 

'* resistance brush 15 91 

" •* brushes. Use of . 13 39 

Holders and rocker. Brush .... 14 58 

Homopolar 13 24 

Hot-wire ammeters and volt- 
meters 17 85 

" •• instruments. Stanley . . 17 86 

Hysteresis loss 12 20 

" 14 17 

" '* Estimation of . . 13 73 

Mag:netic 12 20 

I 
Impedance and reactance of cop- 
per wire 17 81 

Inclined-coil indicating: wattmeter, 

Thomson 17 62 

•* " instruments 17 40 

Indicating: wattmeter. Wagmer . . 17 58 
wattmeter. Thompson 

inclined-coil 17 52 

Iflduced E. M. P., Value of .... 16 33 

Induction, Effects of self 16 29 

g:enerator 19 38 

'* instruments 17 41 

** motor. Characteristic 

curves of 19 51 

*• motor. Power factor of 19 50 
*• ** Wagrner ■ing:le- 

phase .... 19 60 



Digitized by VjOOQ IC 



INDEX 



XV 



Ste. Pagt 

Indnction motors 19 28 

.19 61 

•• " Armatures for 19 84 
Methods of 

startioiT ... 19 41 

SiDfiTle-pbase . 19 52 
•* •* Speed -resfula- 

tion of ... 19 40 

Unipolar 13 24 

Inductor and revolvlnfir-field alter- 
nators 18 25 

Inductors or face conductors ... 12 10 
Instruments. Altematinir-current 

measurinsT .... 17 34 

" Classes of 17 85 

Inclined-coil 17 40 

Induction 17 41 

** Plunsrer and masr- 

netic-vane 17 S9 

Stanley hot-wire . . 17 86 

Insulation of armature conductors 14 12 

Slot 18 61 

" 14 13 

Inverted rotarles 19 71 

n R loss in slots 14 18 

li 

LafiT. Auffle of 16 42 

Lamination 12 19 

Lapwindin^r 18 9 

Leakasre. Coefficient of masmetic 18 82 

Effect of masmetic ... 19 11 

Magnetic 13 82 

19 5 

Lensrth of the armature core ... 14 4 

Line capacity 17 82 

•• Self-induction of 17 29 

Lines, Drop in alternatin^r-current 17 82 

of force, Density of .... 12 44 

" Transmission 17 28 

Load. Non-inductive 16 30 

*• Too much 15 89 

LocatinsT short-circuited armature 

coils 15 99 

Loss. Armature l^ R 14 68 

" Eddy-current 12 18 

13 74 

14 18 

** Estimation of hysteresis . . 18 78 

•• Hysteresis 12 20 

14 17 

" in series field 14 69 

•• •• shunt coils . , 14 69 

•• " slot n R 14 18 

** of residual maametism ... 15 84 

•' Watts 13 76 



Sec. Page 

Losses and heatinsr. Armature . . 13 70 

Armature-core 12 18 

Division of 13 70 

Effect of core 19 12 

Electrical 18 80 

Mechanical 14 69 

Lost, Watts 13 74 

Low bars , 15 79 

** speed . . . . ; 15 87 



M 

Machine. Characteristics of series 
** Characteristics of shunt 
" Dynamo-electric .... 

Machines, Care of 

" Characteristics of com- 

pound-wound .... 

Macmet circuit 

•* " Form of 

•• cores 



12 60 

12 63 

12 1 

15 75 



" " and pole pieces . . 

Desifim of the field . . . . 

" Field 

•• frames 

** wire. Brown ft Sharpe 

sraufire 

Masnetic circuit 



densities 
leaka^re . 



•• " Coefficient of . . 

Effect of . . . . 

•• hysteresis 

** vane and plunsrer instru- 
ments 

*• yoke 

Masmetism. Loss of residual . . . 
Masmetization at pole tips . . . . 

" curve 

Masrnetlzins: current 

Macmets. Multipolar field 

MeasurinsT instruments. Alterna- 

tinff-current 

Mechanical desism 



into electrical enerjry. 
Conversion of . . . 

** losses 

Moisture in field-coils 

Monocyclic alternator 

system 

Motor 

" Action of 

" ** '* shunt 



67 
23 
45 
87 
23 
80 
77 
2 
81 

n 

42 
28 
79 
82 
5 
82 
11 
20 



43 

84 
45 
29 
6 
15 

34 
42 
73 

82 
69 
83 
51 
51 
1 
2 
20 



Digitized by VjOOQ IC 



XVI 



INDEX 



See. Pare 
Motor and dynamo rotation ... 15 82 
** Characteristic carves of in- 
duction 19 61 

** Compound dynamo con- 
verted to 15 36 

•• Counter B. M. P. of .... 16 3 

*• efficiency 16 8 

• Failure of. to start 16 88 

*' Repulsion . 19 68 

" Shunt-wound 15 20 

*• Wagner single-phase indue- 

tion 19 29 

Motors, Accumulatively wound . . 15 81 

" Alternating-current ... 19 23 
and dynamos. Care and 

operation of ...... 15 75 

'* and dynamos compared . 15 1 

" Armatures for induction . 19 34 

Classes of 16 19 

*• Commercial efficiency of 15 10 

•* Compound-wound .... 15 - 80 

•* Design of direct-current . 15 68 

*' Differentially wound . % . 15 80 

•' Direct-current 15 1 

" Induction 19 28 

19 61 

'* Methods of starting induc- 
tion 19 41 

** Operation of direct-cur- 
rent 15 1 

" Power factor of induction 19 50 

Series 15 24 

Shunt 15 20 

" Single-phase induction . . 19 52 
" Speed regulation of induc- 
tion 19 40 

" '* regulation of series 15 28 

" '• regulation of shunt 15 21 

*' Stationary 15 73 

" Synchronous 19 23 

Multipolar field magnets 12 15 

rotary converters ... 19 69 

Multivoltage speed control .... 15 69 

N 

Neutral region 12 12 

Non-inductive load 16 30 

O 

Open-circuited armature 16 91 

field coUs. Testing 

for 16 93 

** coil and closed-coil windings 12 11 

** ** armature windings ... 13 15 
Operation and care of djmamos 

and motors 16 76 



Sec Pure 

Operation from direct-current side 19 76 

" of direct-current motors 16 1 

" *• rotary converters . . 19 71 

Output and power IS 30 

*• current and E. M. P., Re- 
lation between 18 46 

** Determination of 16 69 

•• Estimation of 14 1 

Factors limiting 18 86 

** of constant-potential dyna- 
mos 13 34 

Overcom pounded 12 67 

Overloaded armatures 16 81 

P 
Parallel, Self-induction and ca- 
pacity in 17 7 

'* Resistance and capacity 

In 17 16 

** Resistance, self-induction 

and capacity in .... 17 15 

** winding. Double 13 11 

Single 12 29 

windings 12 26 

13 4 

Period 16 4 

Peripheral speed and diameter . . 14 21 

Pitch of the poles 12 21 

** or spread of coils 12 21 

Plunger and raagnetic-vaoe instru- 
ments 17 39 

Pole pieces 14 23 

*' ** and magnet cores ... 18 80 

" Pluxin 14 28 

** tips. Magnetization at .... 13 45 
Poles. Armature surface covered 

by 14 3 

" Consequent 12 46 

'• Pitch of the 12 21 

•• Salient 12 46 

Polyphase alternators . ' 18 32 

Portable wattmeter 17 49 

Potential transformers 19 21 

Power and E. M. P., Calculation 

of 13 27 

•• •• output 18 . 30 

** ** wattless components . 17 26 

- Calculation of 13 28 

*' expended in a]temating-cm> 
rent circuits. Calculation 

of 17 18 

** factor of a circuit 17 25 

induction motor . 19 50 

Primary and secondary coils. Bfifect 

of resistance of 19 11 

Properties of sine cnnres 16 s 



Digitized by VjOOQ IC 



INDEX 



xvii 



R Sec. Paze 

Radial brashes 15 76 

Radiatjnc: surface 13 75 

Ratio of transformation 19 5 

Reactance 16 36 

16 41 

** and impedance of cop- 
per wire 17 31 

Capacity 16 48 

Reaction. Ampere-turns to offset 

armature 14 36 

" Armature 18 42 

16 17 

Effects of armature . . 13 49 

. . 14 33 

Reentrancy 12 29 

Reentrant windinsT 12 29 

Resrion. Neutral 12 12 

Regulation of induction motors. 

Speed 19 40 

•* " rotary converters. 

Voltagre 19 75 

*• " series motor. 

Speed 15 28 

Reluctance. Control by variation of 

field 15 67 

Repulsion motor 19 58 

Residual maenetism. Loss of ... 15 84 

Resistance 16 41 

and capacity. Circuits 

containins: 17 1 

" and capacity in par- 

aUel 17 5 

" and self-induction. Cir- 
cuits containinsT ... 16 38 
" Circuits containins: . . 16 30 
** Estimation of arma- 
ture 13 70 

'* in armature. Starting 

with 19 46 

'* of prim ary and second- 
ary c.oils. Effect of . 19 11 
*' self-induction, and ca- 
pacity 17 1 

** self-induction, and ca- 
pacity. Circuits con- 
tainins: 17 8 

** self-induction, and ca- 
pacity in parallel . . 17 15 
Resolution and composition of cur- 
rents and E. M. P.'s 16 20 

Resonance. Electrical 17 13 

Reversed or short-circuited arma- 
ture coil 15 91 

ReversinsT direction of rotation . . 15 46 
** switch. Shunt motor 

with 15 49 



Sec. Pare 

Reversing: switches 15 47 

Revolutions per minute. Speed in . 14 7 

Revolvins: field 19 30 

** ** and inductor alter- 
nators 18 25 

Rheostat startinsr 15 35 

Rheostatic control 15 22 

Rheostats, Automatic starting: . . 15 56 

" with automatic release 15 38 

Rlns: and drum armatures .... 12 10 

" windings 13 1 

Rings. Collector 12 5 

Cross-connecting 14 20 

Rocker and holder. Brash 14 58 

arm 12 18 

Rotaries. Inverted 19 71 

Rotary converters 19 64 

Heating of. . . 19 64 
Multipolar ... 19 69 
Operation of . 19 71 
*• " Voltage regula- 
tion of ... . 19 75 

'* transformers 19 (54 

Rotation, Dynamo and motor . . L*) 32 

** Reversing direction of . 15 46 

.. jg gg 

Speed and direction of . 19 28 

Rotor ". . 19 37 

S 

Salient poles 12 46 

Secondary and primary coils. Ef- 
fect of resistance of 19 11 

Segments. Number of coi^mutator 14 10 

Self-excited dynamo 12 3 

" induction and capacity .... 16 28 
** " and capacity. Cir- 
cuits containing . 17 6 
** " and capacity in par- 
allel 17 7 

•' *' and resistance. Cir- 
cuits containing . 16 38 
•* " Circuits containing . 16 31 

of line 17 29 

** ** resistance and ca- 
pacity 17 1 

•* " resistance and ca- 
pacity. Circuits 
containing .... 17 8 
** *• resistance and ca- 
pacity in parallel 17 15 

Separately-excited dynamo 12 3 

Series dynamo converted to series 

motor 15 84 

** field. Loss in 14 69 

" ** winding IS 85 



Digitized by VjOOQ IC 



xvin 



INDEX 



Sec. Pare 

Series machine. Characteristics of 12 60 
motor across constant-po- 
tential mains ... 15 27 
" '* connections .... 15 60 
Desifirn of 10-horse- 

power 15 72 

'* " on altematingr- cur- 
rent 19 57 

" ** ** constant-current 

circuit .... 15 29 
** " constant -poten- 
tial circuit . . 15 24 
*' " Series dynamo con- 
verted to 15 S4 

" '* Speed ree^ulation of 15 28 

motors 15 24 

*• transformers ........ 19 20 

•• winding 12 58 

14 38 

Double 13 11 

" " Singrle 12 S5 

Triple 12 40 

•• windin^rs 12 33 

13 9 

Double 12 39 

Shaft and spider . 14 43 

Shafts 13 62 

Dynamo 13 62 

Shell transformers 19 14 

Short-circuited armature coils. 

Locatinir .... 15 99 
field coil ..... 15 94 
" " or reversed arma- 
ture coil .... 15 91 

•' circuits 15 86 

infield 15 88 

Shunt coils, Loss in 14 69 

Methods of connecting: 15 40 
" dynamo converted to shunt 

motor 15 S2 

*' field, Wrpns: connections of 16 43 

•• •• 15 88 

machine. Characteristics of 12 63 

motor. Action of 15 20 

" connections 15 37 

" " Design of lO-horse- 

power 15 71 

" ** on altematinsr cur- 
rent 19 57 

" " Shunt dynamo con- 
verted to 15 32 

*• " with reversintr 

switch 15 49 

•• motors 15 20 

'* ** Speed regulation of 15 21 

•• windinsr 12 62 



Sec. 

Shunt wlndinsr 14 

Determination of . 13 

** wound motor 16 

Sine curves 16 

Addition of 16 

" Properties of 16 

*' waves. Values of 16 

Sinsrle-layer windinar 13 

" parallel winding 12 

** phase alternators 18 

18 

** converters 19 

" " induction motors . . 19 

*' induction motors. 

Wagmer 19 

*' series winding 12 

** windinsr 12 

Sinsrly reentrant, double, parallel 

windinff ... 12 
" ** double windinsr 12 
Size and arransrement of con- 
ductor 14 

Slip 19 

" and torque. Relation between 19 

Slot, Dimensions of 14 

" insulation 13 

14 

Slots and coils. Number of .... 14 

" Armature 13 

" /«^lossin 14 

Smooth -core and tooth armatures 13 

Sparkinc: 15 

*' and commutation .... 13 
Sparkless commutation. Require- 
ments for 13 

Speed and direction of rotation . . 19 

** ** torque curves 16 

" control. Multivoltatre ... 15 

" in revolutions per minute . 14 

" Low 15 

" resfulation of induction 

motors .... 19 

" " of series motors 15 

" too hisrh 16 

Spider and core. Construction of . 18 

" shaft 14 

Spiders. Armature 13 

Spread or patch of coils ...... 12 

Squirrel-caare armature 19 

Stanley electrostatic voltmeter . . 17 

hot-wire instruments ... 17 

" starting: compensator . . 19 

Star and delta connections .... 18 

Start, Failure of motor to 16 

Starting: compensator, or auto- 
transformer . ; 19 



Paze 

89 

86 

20 

6 

9 

9 

22 

8 

29 

1 

32 
64 
52 

69 
35 
29 

32 
81 

20 
36 
87 
13 
61 
13 
11 
58 
18 
57 



41 
28 
26 
59 
7 
87 

40 
28 
90 
50 
43 
62 
21 
85 
55 
36 
41 
42 
88 

41 



Digitized by VjOOQ IC 



INDEX 



XIX 



Sec. Page 

Startinsr compensator, Stanley . 19 41 
" induction motors. 

Methods of 19 41 

rheostat 15 35 

** ** Automatic ... 15 56 
'* Automatic- 
release .... 15 44 
'* with resistance in arma- 
ture 19 46 

Stationary motors 15 73 

Stator 19 .37 

Switch. Shunt motor with revers- 

inir . . . . .' 15 49 

Switches, Reversinsr 15 47 

Synchronous converters 19 64 

motors 19^ 23 

System, Monocyclic 18 51 

'* Two-phase and three- 
phase 16 18 

T 

Table of alternators 18 12 

. ** ** armature flux 14 81 

" " commercial efficiency of 

motors 15 10 

" ** dynamo windings .... 12 41 
'* ** electrical efficiencies of 

dynamos 14 2 

" " induction motors .... 19 61 
" " masrnet wire. Brown & 

Sharpe crausre 18 72 

** masmetic circuit 14 28 

" " reactance and impedance 

of copper wire 17 31 

" values of K in formula 12 13 63 

" watts loss 13 76 

lost 13 74 

Tansrential brushes 15 76 

Teaser system of control 15 .66 

Teeth, Armature 13 80 

Density hi 14 14 

Temperature, Rise in 14 73 

Test. Bar-to-bar 15 79 

** for defects in armature ... 15. 97 
" sfrounds between wind- 
ins: and frame 15 95 

TestiuK 14 72 

for faults 15 92 

open - circuited field 

coils 15 93 

Theory of the dynamo 12 1 

Thomson-Houston constant -cur- 
rent dynamo 13 19 

" Inclined - coll indicating: 

wattmeter 17 52 

Three-phase alternators 18 89 



Su. Page 
Three-phase and two-phase sys- 
tems 16 18 

" *' compound-wound 
alternators, West- 

in8:house 18 55 

" " converters 19 66 

Tooth and smooth-core armatures 13 57 

Toothed armatures 12 18 

Torque 15 11 

and current. Relation be- 
tween 15 16 

** " slip. Relation between 19 37 

** '* speed curves 15 26 

Transformation, Ratio of 19 5 

Transformer. Action of ideal ... 19 8 

Theory of the ... 19 8 

Transformers 19 1 

Construction of . . 19 14 

Core 19 14 

" Examples of ... 19 15 

Potential 19 21 

Rotary ........ 19 64 

Shell 19 14 

Transmission lines 17 28 

Triple series windinirs 12 40 

windings 12 38 

Two-layer windings 13 8 

** phase alternators ...... 18 32^ 

** ** and three-phase sys- 
tems .16 18 

'* " compound -wound 
alternators, Westin^r- 

house 18 58 

** converters 19 65 

U 

Unicoil windintrs 18 37 

Unipolar dynamos 13 24 

induction . . 13 24 

V 

Value of induced B. M. P 16 33 

Valuesof /Tin formula 12 13 63 

** sine waves 16 22 

** Relations between 16 26 

Variation of field reluctance. Con- 
trol by 15 67 

Ventilating: ducts 13 57 

Vibration . 15 91 

Voltatre retrulation of rotary con- 
verters 19 75 ■ 

Voltmeter. Stanley electrostatic . 17 55 
** Wagner altemating:- 

current 27 46 

** Weston altemating:- 

current 17 45 



Digitized by VjOOQ IC 



INDEX 



Sec. Page 
Voltmeters and ammeters. Hot- 
wire 17 85 

Electrostatic 17 M 

Volts, Calculation of field winding 

for 115-126 14 87 

•• Windins: for 125 14 20 

"260 14 19 

W 

Warmer alternatinsr-current Tolt- 

meter 17 46 

indicatins: wattmeter. . . 17 58 
*' sinirle-phase induction 

motor 19 69 

Wattless and power components . 17 26 

Wattmeter. Compensated 17 60 

Portable 17 49 

" Thomson inclined- 
coil indicatins: ... 17 52 
Wasrner indicating: . . 17 58 
'* Weston compensated 17 61 

Wattmeters 17 48 

Watts loss 13 76 

••' lost 13 74 

Wave forms. E. M. F 16 1 

winding 13 9 

Waves. Value of sine 16 22 

Weak field 15 92 

Wcstinsrhouse three-phase com- 
pound-wound al- 
ternator 18 66 

" two-phase com- 
pound-wound al- 
ternator 18 58 

Weston alternating -current volt- 
meter 17 46 

•* compensated wattmeter 17 51 
** portable wattmeter ... 17 49 
Windage and friction. Bearing . . 14 68 
Winding and core, Design of arma- 
ture 14 1 

" Compound 12 66 

** Determination of series- 
field 13 85 
" ** *' shunt 13 86 

Double 12 80 

•• paraUel 18 11 



Sec. Page 

Winding. Double series 18 11 

for 125 volts 14 20 

" 250 volts ....... 14 19 

Lap 18 9 

Reentrant 12 29 

requirements 12 87 

Series 12 68 

14 38 

Shunt 12 62 

14 39 

Single 12 29 

** parallel 12 29 

Singly reentrant double 

parallel 12 82 

Style of 14 20 

the coils. Manner of ... 12 21 

Two-layer 12 23 

Wave 18 9 

Windhigs, Closed-coU 18 1 

- ** armature . 12 20 

Computation of field . . 14 29 

Double series 12 39 

Drum 12 10 

18 4 

Dynamo 12 41 

Field 13 84 

Method of applying . . 18 60 
Open-coil and closed- 
coil .... 12 11 
" " armature . . 18 15 " 

ParaUel 12 26 

18 4 

Ring 18 1 

Series 12 83 

... 18 9 

Single-layer 13 8 

** series 12 35 

Triple 12 33 

series 12 40 

Two-layer 13 8 

Unicoil 18 87 

Wire. Magnet. Brown & Sharpe 



gauge 



13 72 



Yoke 13 80 

" 14 26 

*' Magnetic 12 43 



Digitized by VjOOQIC 



> 



Digitized by VjOOQIC 



5|it| 



168 »/" 

^ 70019 4^-ill 




Digitized by VjOOQIC 



"Vjr 



Digitized by VjOOQIC 



Digitized by VjOOQIC 



\. 



Digitized by VjOOQIC 



a«i07ni72Sb 




b89071917256a 



Digitized by VjOOQIC 



4 



DCMOO 



K.F. WENDT LIBRARY 

UW COLLEGE OF ENGR. 

215 N.RANDALL AVENUE 

MADISON 




"-D ipfeecfby' Lj O OQ IC 



Digitized by VjOOQIC